Journal of Retailing and Consumer Services 74 (2023) 103415

Contents lists available at ScienceDirect

Journal of Retailing and Consumer Services

journal homepage: www.elsevier.com/locate/jretconser

Images, reviews, and purchase intention on social commerce: The role of

mental imagery vividness, cognitive and affective social presence
Erik Ernesto Vazquez a, *, Chirag Patel b, Salvador Alvidrez c, Lorena Siliceo d
a
Léonard De Vinci Pôle Universitaire, Research Center, 92916, Paris La Défense, France
b
Grenoble École de Management, Marketing Department, 38000, Grenoble, France
c
Queen’s University, Psychology Department, Belfast, BT9 5BN, United Kingdom
d
Universidad Anáhuac México, Facultad de Economía y Negocios, México

A R T I C L E I N F O A B S T R A C T

Handling Editor: Prof. H. Timmermans Images and textual reviews are central to social commerce sites. But little is known about the mechanism by
which image quality and type of textual review (with personal experiences versus facts) impact consumer pur­
Keywords: chase intention on a social commerce page. We address this knowledge gap and use the stimulus-organism-
Mental imagery vividness response (S–O-R) framework to develop our conceptual framework. We hypothesize that image quality and
Cognitive social presence
type of textual review impact purchase intention through imagery vividness that functions as a stimulus. Further,
Affective social presence
imagery vividness impacts purchase intention positively in the social commerce context through social presence.
Social commerce
Specifically, the effect of imagery vividness on purchase intention is likely to be serially mediated by cognitive
social presence and affective social presence. We collect data by conducting two experiments using a social media
post regarding fashion apparel for middle-aged women on Instagram and find support for our thesis.

1. Introduction role that social influence plays in this context. This approach will allow
us to study purchase intention as a socially influenced decision, instead
The use of social media activities to support b-to-c and c-to-c e- of a predominantly individual decision, as it is the case of e-commerce
commerce transactions is referred to as social commerce (Liang and platforms (see Kim, 2015).
Turban, 2011-12; Wu and Li, 2018). This includes social networking Research on social commerce comprises a large stream of studies that
sites that incorporate commercial features to allow transactions. Face­ seek to understand the factors that impact consumer behavior. Studies
book alone had 60 million active business pages that supported e-com­ on participation/purchasing intention and buying in the social com­
merce transactions in 2016 (venturebeat.com). Instagram has more than merce context have identified drivers such as relationship quality (Liang
25 million business pages, and more than 80% of Instagram users follow et al., 2011), trust towards members (Farivar et al., 2017; Wu et al.,
at least one business (Wham, 2018). Social commerce involves con­ 2023), social presence (Kim, 2015; Nadeem et al., 2020; Zhang et al.,
necting customers to discover, share, review, recommend, rate products, 2014), social support (e.g. Liang et al., 2011; Zhang et al., 2014),
and initiate or simplify purchase decisions (Olbrich and Holsing, perceived values like utilitarian, hedonic, and social values (Gan and
2011-12). Product images uploaded by sellers -firms or individuals- and Wang, 2017), reciprocity (Leung et al., 2020), and satisfaction (Zhang
textual reviews contributed by individuals from the social network and Luo, 2016). Research on the role of images and reviews as drivers of
comprise the bulk of the content on social commerce. The use of images consumer behavior on social commerce is limited and is elaborated
in social commerce is crucial and constitutes a growing portion of the below.
billions of images shared daily on popular media platforms. In 2015, 3.2 Images play a crucial role in impacting user engagement on social
billion images were shared daily on popular social media platforms by media platforms (Sabate et al., 2014; Li and Xie, 2020) and online
individuals and firms (Meeker, 2016). Since product images and text purchasing decisions in e-commerce (Chau et al., 2000; Li et al., 2014).
reviews on social media represent the two essential elements of social For the social commerce context, the research on images is sparse so we
commerce (Han et al., 2018; Huang and Benyoucef, 2017), we seek to know little about the performance effects of images on transactional
understand how these contributions impact purchase intention and the consumer response in the social commerce context. The results of the

* Corresponding author.
E-mail address: [email protected] (E.E. Vazquez).

https://doi.org/10.1016/j.jretconser.2023.103415
Received 3 February 2023; Received in revised form 5 April 2023; Accepted 1 May 2023
Available online 2 June 2023
0969-6989/© 2023 Published by Elsevier Ltd.
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

studies in the e-commerce context cannot be generalized to social in studies of social commerce (Friedrich et al., 2019; Wu et al., 2023;
commerce due to certain characteristics of social commerce (Busalim Zhao et al., 2023), influence the effect of imagery vividness on purchase
and Hussin, 2016). First, social commerce allows images to be uploaded intention, we include cognitive and affective dimensions of social
by users that are non-professionals resulting in substantial heterogeneity presence as serial mediators in our framework. Thus, this study also
in the quality of images uploaded. Second, social influence is a crucial contributes to weighing the relevance of the “awareness” of other con­
aspect of social commerce that needs to be considered to better under­ sumers in social commerce platforms compared to the extent to which
stand the factors influencing purchase intention in this context (Xu et al., others’ reviews influence the consumer’s impressions.
2019). Thus, despite substantial research on the performance effects of To empirically examine the proposed effects of image quality and
images on social media and e-commerce, little is known about the nature review type, we conducted two experiments using a mock social media
and mechanism through which images impact social commerce. post regarding a specific product category on Instagram, a leading social
Customer reviews on social commerce platforms can impact user’s networking platform for sharing images. We empirically examine the
purchase behavior through social influence (Xu et al., 2019; Cao et al., proposed effects of image and textual reviews in fashion apparel for
2020, 2021). Quality, valence and linguistic aspects of reviews have middle-aged women. This consumer segment, with great influence and
been shown to influence purchase intentions in the social commerce veto power over retail purchase decisions, drives 70%–80% of all con­
context (Li, 2019; De Vries et al., 2012; Lee and Shin, 2014; Wang et al., sumer purchases (Berg and Amed, 2020; Brennan, 2015; Statista, 2019).
2023). In contrast, in this paper we focus on review content in terms of Beyond these economic reasons, this segment is also relevant because
whether they are based on personal experiences or on a factual list of the receptivity of marketing messages and mental imagery processes are
attributes. For example, a product review based on a personal experi­ also moderated by gender and age (Kim et al., 2021; Taylor et al., 2011),
ence would be: “My girlfriend LOVES her Tweed jacket. When I told her which is one of the reasons for which fashion apparel brands extensively
she could wear it under the rain, she was amazed …“. In contrast, a use social media to engage with this segment (Kefi and Maar, 2018; Park
fact-based product review would be: “This Tweed jacket can be a great et al., 2020).
gift. You can wear it under the rain and enjoy all the features …“. Since Accordingly, the contribution of the present study to the extant
extant literature does not study review content type or image quality in literature is two-fold: first, our research contributes to the rapidly
social commerce platforms and its potential effect on their sales, we growing literature on the content effectiveness of social commerce.
address this knowledge gap. Specifically, we seek to understand the Given the rapid rise of social commerce platforms, it is timely that we
mechanism by which image quality and the type of textual review (personal expand the current knowledge on social commerce content effectiveness
experiences versus facts) impact consumer purchase intention in a social by exploring the mechanisms by which image content and review con­
commerce page. tent affect purchase intention. Therefore, our paper offers essential
In this study, we aim to answer these questions by examining the managerial implications to social commerce content creators, particu­
impact of image quality and the type of textual review on purchase larly about enhancing content effectiveness, managing image quality,
intention. Based on the concepts of mental imagery vividness and social and influencing review content. These insights are especially useful in a
presence, we propose two mechanisms by which image quality and context in which customers increasingly seek to influence their social
textual reviews impact purchase intention. Mental imagery vividness is networks by posting purchase reviews.
one of the dimensions of visual imagery,1 and it is described in terms of Second, our research also contributes to the visual marketing aspect
the clarity and liveliness of images (Marks, 1973; Bone and Ellen, 1992; of social commerce literature. Li and Xie (2020) point out that extant
Miller et al., 2000; Andrade et al., 2014). Imagery vividness has been a research has primarily focused on the impact of images on advertising
strong psychological driver of the consumer’s response in online envi­ effectiveness (e.g., Fortin and Dholakia, 2005). However, although
ronments (e.g., Ching et al., 2013; Vazquez, 2020). Likewise, social individually based purchase intention has worked well as a measure of
presence is an individual’s sense of awareness of the presence of others advertising effectiveness in traditional offline and online channels, the
in online settings (Short et al., 1976; Lu et al., 2016). It also has been a social commerce context requires scholars to consider purchase inten­
prominent concept in computer-mediated communication that seeks to tion as a socially influenced decision. Our research extends the literature
explain the capacity of transmitting personal and sociable messages in on visual marketing by considering the social influence of reviews.
online environments (Guo et al., 2022; Nadeem et al., 2020; Weisberg, Particularly, we assess cognitive and affective social presence in-terms
Te’eni, and Arman, 2011). In our conceptual framework (see Fig. 1), of the social influence of reviews and consider them as key elements
mental imagery vividness works as an external stimulus that mediates of the mechanism that links mental imagery vividness to the consumers’
the impact of image quality and the type of textual reviews on purchase purchase intentions in social commerce.
intention. In this context, imagery vividness is also expected to influence
perceptions of social presence, which operates as a consumers’ internal 2. Background
state. In addition, in order to explore the extent to which cognitive and
affective states, which are the two types of consumer reactions identified In this section, we review related literature and discuss how our
study builds on and extends the existing literature. Table 1 provides a
summary of this extant literature which we critically review and build
upon in-order to propose a conceptual modeling framework for the de­
terminants of purchase intention in social commerce.

2.1. Theoretical framework

In this study, we use the stimulus-organism-response (S–O-R) theo­

Fig. 1. Conceptual framework. retical framework to understand the mechanism by which stimuli like
images and reviews impact consumer purchase intention in social
commerce. The S–O-R framework from environmental psychology
(Mehrabian and Russell, 1974) was developed to understand the shop­
ping environment in-terms of stimuli that affect the organisms’ (con­
1
The other two dimensions of visual imagery are valence (i.e., the goodness sumers’) internal state and their external response behaviors towards
or badness of visual imagery), and quantity (i.e., the number of visual images physical stores (Bitner, 1992; Spangenberg et al., 1996; Slama and
evoked by a stimulus). Tashchian, 1987; for a review see Turley and Milliman, 2000). This

2
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

Table 1
Literature review of studies examining images and textual reviews in social media, e-commerce, and social commerce.
Author(s)/ Research focus Independent variables Dependent variables Results
Publication
date

Cao et al. To investigate effect of user proximity and Customer reviews Conversion rate on Social influence from users with high
(2020) loyalty on purchasing behavior in social social commerce proximity explains high conversion rate on
commerce. demonstrates sig- nificant Beidian, a social commerce site.
property deviations from all known social
networks and E-Commerce in terms of
network structure, dynamics and user
behavior.
Cao et al. To understand why users engagement in Customer reviews Purchases on social Role of reviews and recommendations from
(2021) Instant messaging (IM)-based social commerce pre-existing relations is investigated through
commerce. qualitative interviews to understand the
benefits to users of IM-based social
commerce.
Chau et al. Impact of product information Impact of images versus text Purchase intention on For familiar products, images are better than
(2000) presentation modes on Online shopping. e-commerce sites text in terms of both efficiency and
effectiveness. When users are not familiar
with the product, the advantages of images
over text diminish.
Chevalier and The effect of consumer reviews on online Online reviews Consumers’ attitudes Improvement in review valence leads to an
Mayzlin purchase and decisions on e- increase in relative sales in Amazon.com and
(2006) commerce sites BarnesandNoble.com. The impact of 1-star
reviews is greater than the impact of 5-star
reviews.
Chintagunta The impact of valence, volume, and Online reviews Consumers’ attitudes Valence of reviews has a greater effect on
et al. (2010) variance of online user reviews on box and decisions in e- movie box office performance than the
office performance of movies. commerce volume of reviews.
Coyle and The effect of progressive levels of Image quality, vividness Attitude towards online Increases in interactivity and vividness are
Thorson interactivity and vividness on attitude shopping Websites associated with increased feelings of
(2001) toward web sites, feelings of telepresence, telepresence. Also, increases in vividness are
and attitude-behavior consistency. associated with more positive and more
enduring attitudes toward the Website.
Cox and Park To determine which features of Images Attracting consumers’ An orthodontist’s credentials, before-and-
(2014) orthodontists’ Facebook pages are the attention on social after photographs, consumer-focused
most effective in attracting patients. media content, and the orthodontist’s photograph
are key elements of Facebook that attract
attention of Facebook users and aid decision
making about where to receive treatment.
Fortin and The effects of various levels of interactivity Vividness Attitude and behavioral Moderate effects of interactivity and
Dholakia and vividness of a message on attitudes intentions on e- vividness on social presence. The effects of
(2005) and behavioral intentions within a web- commerce sites interactivity reach a ‘‘plateau’’ at medium
based advertisement. and high levels, indicating a diminishing
returns effect. The impact of vividness is
linear with a steady increase across low,
medium, and high levels.
Gavilan et al. Effect of mobile advertising and the Vividness Trust and purchase MMS mobile ads and transformational ads
(2014) mediating effect of mental imagery intention on e- have greater impact on vividness and
elicited by mobile advertising on trust in commerce sites elaboration, while SMS mobile ads have a
the advertising message and purchase greater impact on the quantity dimension.
intention. Mental imagery elicited by mobile ads
increases ad trust. Vivid and elaborate mental
imagery mediates the effect of the type of ad
on ad trust and has a positive influence on
purchase intention.
Hassanein and The effect of social presence on Images Consumers’ enjoyment The various levels of socially rich text and
Head (2007) consumers’ attitude towards online and trust towards picture design elements impact perception of
shopping. online shopping online social presence. Higher levels of social
presence positively impact the perceived
usefulness, trust, and enjoyment on shopping
Websites.
Hou and Pan The impact of hotel photo aesthetics on Image aesthetics, such as color User engagement on Aesthetically enjoyable photos encourage
(2023) consumer engagement. harmony, vivid color, lighting, depth social media more consumers to engage in word-of-mouth
of field, and content discussions and promote consumer ratings.
This impact is moderated by hotel price.
Jiang et al. To identify the key convenience Textual information Consumers’ attitudes The five dimensions of online shopping
(2013) dimensions of online shopping that impact and decisions in e- convenience are: access, search, evaluation,
adoption of online shopping. commerce transaction, and possession/post-purchase
convenience. Evaluation convenience
includes use of text and graphics for
providing product information.
Kang et al. To examine: first, whether the consumer Individual consumer characteristics Purchase intention on Among the CSI consumer decision-making
(2014) style inventory (CSI) consumer decision- like novelty, brand, and price social commerce styles, novelty/fashion consciousness is the
making styles were related electronic word consciousness most important antecedent of opinion
of mouth (eWOM) in social media and seeking using eWOM. The brand
(continued on next page)

3
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

Table 1 (continued )
Author(s)/ Research focus Independent variables Dependent variables Results
Publication
date

attitudes toward online social shopping; consciousness decision-making style is the

and second, whether opinion seeking in most important antecedent of favorable
social media and attitudes mediated the attitudes. Novelty/fashion consciousness,
links between decision-making styles and brand consciousness, and price consciousness
purchase intent. decision-making styles have indirect effects
on social shopping, mediated by both opinion
seeking and favorable attitudes.
Kim (2015) The effect of vividness, interactivity, and Mental imagery vividness, presence Participation intention Social presence and telepresence mediate the
presence on participation intention. on social commerce effect of interactivity and vividness on
sites usefulness and enjoyment. Both usefulness
and enjoyment significantly influence
consumer participation intention on social
commerce.
Senthil Kumar To develop a measure for para-social Need for online social connection Purchase intention on Develops a research framework that
et al. (2015) presence. social commerce illustrates the impact of new technologies
(such as personalization systems) and
associated web interface design decisions on
perceived communication characteristics of a
web site, para-social presence, and
subsequent user evaluations of the web site.
Lee and Shin Effect of review quality, product type, and High quality reviews Consumers’ attitudes High quality reviews lead to more positive
(2014) reviewer’s photo on online purchase and decisions in e- evaluations than low quality reviews and this
intention. commerce leads to stronger purchase intention. Review
quality also has a negative direct effect on the
purchase intention for the experience good,
with no corresponding effect for the search
good.
Lee et al. The effect of Facebook users’ arousal and Arousal and valence in users Purchase intention on A user’s emotional connection to a Facebook
(2012) valence on intention to participate in an social commerce event page increases the event page’s
event. perceived usefulness and ease of use, as well
as the user’s perceived enjoyment, which
could lead the individual to attend the actual
event. Furthermore, this study shows that
arousal significantly stimulates the valence of
users.
Li (2019) To investigate the influences of social Ratings, valence of reviews, Purchase intention on Social commerce constructs exert positive
commerce sites on customers’ virtual recommendations, and referrals social commerce effects on cognitive states i.e., social
experiences and their purchase intention. presence, informational support, and
emotional support, and affective states i.e.
familiarity and closeness. Furthermore, social
presence and informational support influence
trust in product recommendations, which
impacts social shopping intention.
Li et al. (2014) The impact of product image Image characteristics such as amount Purchase intention on Online consumers prefer product images with
characteristics on purchase decisions. of information embedded, e-commerce sites a larger key object, lower entropy on key
colorfulness, presence of human along objects, a warmer color, a higher contrast, a
with aesthetics like contrast, depth higher depth-of-field, and more social
presence.
Li and Xie The influence of image content (image Image characteristics, such as image User engagement on A positive effect of image inclusion on user
(2020) inclusion and image characteristics) on quality, image source, colorfulness, social media engagement. High-quality and professionally
social media engagement. and presence of human face shot images lead to higher user engagement.
The effect of colorfulness varies by product
category, while the presence of human face
and image–text fit varies by social media
platform used.
Limayem et al. The impact of different factors on online Textual product information Consumers’ attitudes Site accessibility, site loading speed, product
(2000) shopping intentions and behavior. and decisions in e- description, and transaction efficiency were
commerce identified as factors that have a positive effect
on attitude and online purchase intention.
Ma and Meng The impact of image position and layout Image position in tweets with multiple User engagement on Position effects are positive and became
(2021) on user engagement on social media. images social media symmetric with image adding. Layouts with
2, 3, 4, 5, 6, 8 images have positive effects on
user engagement, while layouts with 7 and 9
or more images have negative effects on the
Weibo platform.
Sabate et al. The influence of the content’s richness and Image (inclusion vs exclusion) User engagement on The richness of the content (inclusions of
(2014) time frame on the number of comments social media images and videos) raises the impact of the
and likes in Facebook fan pages. post in terms of likes. On the other hand,
using images and a proper publication time
are significantly influencing the number of
comments, whereas the use of links may
decrease this metric.
(continued on next page)

4
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

Table 1 (continued )
Author(s)/ Research focus Independent variables Dependent variables Results
Publication
date

Song et al. The effect of destination image themes on Image themes User engagement on Image themes have significant and distinct
(2021) social media user engagement in Facebook social media. effects on three indicators of social media
brand pages engagement: the numbers of likes, comments,
and shares.
Song and Kim The effect of online product presentation Image size Purchase intention on Online product presentations influence
(2012) methods on perceived risk of online e-commerce sites consumers’ mental intangibility.
shopping.
Wang et al. To identify relevant linguistic features of Linguistic features of reviews Purchase intention on The number of customers who make a
(2023) marketing content and understand how social commerce purchase increases 1) at the micro-level due
they impact consumer purchase behaviors to self-referencing and detailing, 2) at the
in social commerce. macro-level due to argument structuring, and
3) at the meta-level due to linguistic style
matching, while linguistic content matching
negatively affects the number of customers.
Wu and Li To explore the effects of six marketing-mix Marketing mix variables Customer loyalty on All components of social commerce
(2018) components on consumer loyalty through social commerce marketing mix (SCMM) have significant
consumer value in social commerce. effects on consumer value, which positively
influences customer loyalty in social
commerce.
Xu et al. (2019) To examine how social factors affect user’s Customer reviews User’s purchase Building on social influence of reviews and
purchase behavior in social commerce behavior on social word-of-mouth, the authors propose four
context. commerce platforms primary mechanisms for efficient purchase
conversion: better matching, social
enrichment, social proof and price sensitivity.
Yuan et al., To study persuasive tactics employed in an Customer reviews within online Attractiveness, A multi-task learning prediction model is
(2021) advertising text and predict the online advertisements conversion rate and proposed and evaluated. This prediction
promotion effectiveness. sales volume in social model effectively recognizes the persuasive
commerce tactics employed in an advertising text and
predicts their promotion effectiveness.

Note: The independent variables in the table only include those relevant to the present research.

framework has been applied extensively to e-commerce (see Eroglu internal state to customer loyalty on social commerce. Li (2019) studies
et al., 2001, 2003; Wang et al., 2010; Peng and Kim, 2014; Richard, impact of ratings, valence of reviews, recommendations, and referrals
2005; Sautter et al., 2004; Mollen and Wilson, 2010) and more recently through sequential effect of social interactions and trust as internal
to the social commerce context (Kim, 2015; Li, 2019; Wu and Li, 2018). states on social shopping intention. Although none of these papers study
In the section below we review the existing literature on consumer images or reviews (content aspect) as stimuli, there is substantial focus
purchase intention in social commerce and explore the link between on understanding social factors as consumers’ internal state that impacts
images, reviews, and purchase intention in this context. purchase intention as a consumer response in social commerce.

2.1.1. Consumer purchase intention on social commerce 2.1.2. Impact of images on social commerce
A large body of research on social commerce content suggests that Images can significantly impact user engagement on social media
consumer purchase can be affected by various factors. Zhang and platforms. Studies show that image inclusion (versus exclusion) (Sabate
Benyoucef (2016) provide a thorough review of drivers of different et al., 2014); image characteristics, such as image quality, image source,
consumer responses, from attention and information seeking to pur­ colorfulness, and presence of human face (Li and Xie, 2020); and image
chase intention, purchase behavior, and repurchase in the social com­ aesthetics, such as color harmony, vivid color, lighting, depth of field,
merce context. The authors classified the extant literature using the and content (Hou and Pan, 2023), all have an impact on user engage­
S–O-R framework. Accordingly, we identified the following stimuli as ment on social media platforms. Additionally, image position in tweets
impacting purchase intention on social commerce: individual consumer with multiple images (Ma and Meng, 2021); as well as image themes
characteristics like novelty, brand, and price consciousness (Kang et al., (Song et al., 2021) are also relevant. Images are also useful for impacting
2014); need for online social connection (Senthil Kumar et al., 2015), other performance outcomes like attracting consumers’ attention on
and arousal and valence in users (Lee et al., 2012). The internal state social media (Cox and Park, 2014). Despite substantial research interest
factors identified in these studies are opinion seeking using e-word of in understanding the performance effects of images on social media,
mouth (eWOM) (Kang et al., 2014), and perceived usefulness, ease of little is known about the underlying mechanism.
use, and enjoyment (Lee et al., 2012). In a more recent review of social Images also play a crucial role in e-commerce and significantly
commerce literature, Mou and Benyoucef (2021) found that extant impact online purchasing decisions. Extant literature has examined
research focused on investigating the antecedents of consumer behavior impact of images (versus text) (Chau et al., 2000); image characteristics
such as engagement, buying, purchase intention, e-WOM and social such as amount of information embedded, colorfulness, presence of
sharing intention (Mou and Benyoucef, 2021), They classified the an­ human along with aesthetics like contrast, depth, (Li et al., 2014); and
tecedents covered in these studies into social, technology-related, image size (Song and Kim, 2012) on purchase intention in the e-com­
trust-related, negative valence and positive valence factors. merce context. Images have been shown to impact other performance
Furthermore, there is a growing trend to use the S–O-R theoretical outcomes like consumers’ enjoyment and trust towards online shopping
framework and focus on understanding purchase intention as a con­ (Hassanein and Head, 2007). For the social commerce context, the
sumer response. Kim (2015) links interactivity and vividness as stimuli research on images is limited and as discussed earlier, the results of the
through social presence, telepresence, cognition and affect (internal above studies on e-commerce cannot be directly generalized to the social
states of a consumer) to purchase intention. Wu and Li (2018) link commerce context.
marketing mix variables through customer value as a consumer’s We use media richness theory to understand the mechanism by

5
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

which image quality impacts purchase intention on social commerce with factual list of attributes, on social commerce impacts perceived
platforms. We focus on image quality because among the various image mental imagery vividness positively.
attributes and aesthetics studied, image quality and image source
(professional versus amateur photographer) were found to have positive 2.1.4. Impact of mental imagery vividness on consumer purchase intention
and significant impact on user engagement across two product cate­ on social commerce: The role of social presence
gories and social platforms as compared to image aesthetics variables In their seminal piece on the social psychology of telecommunica­
that were specific to the travel sector (Li and Xie, 2020; Hou and Pan, tions, Short et al. (1976) introduce the concept of social presence.
2023). Consistent with media richness theory (Walther, 1992) we view Originally referred to “the degree of salience of the other person in the
images as rich media tools that enhance mental imagery vividness, with interaction” (p. 64), social presence has come to be used as a parameter
image quality used frequently in experiments to manipulate mental for the involvement of users with other users in online interactions.
imagery vividness (Coyle and Thorson, 2001). Vividness is an important Nowak and Biocca (2003) distinguish among telepresence (a medium’s
dimension of imagery which is a multidimensional construct, encom­ capability of making users feel as if they are ‘inside’ the virtual envi­
passing vividness, quantity, ease, and ability to manipulate images ronment), co-presence (the feeling that the user is in the company of
(Bone and Ellen, 1992). Mental imagery vividness is more dependent on other users), and social presence (the medium’s capability of conveying
the quality features of images (sharpness, resolution, brightness) than access to another mind). As such, social presence has gone from being a
imagery quantity. Previous empirical studies in digital marketing have medium’s feature to becoming a psychological variable mirroring a
demonstrated positive, significant effects emerging from the vividness sense of connectedness in online interactions, which conversely affects
dimension of imagery (depth or quality), in contrast to poor effects users’ attitudes and behaviors. Gefen and Straub (2004) argue that so­
observed from the breadth or quantity dimension of imagery (Fortin and cial presence can even be perceived in absence of other humans; they
Dholakia, 2005; Gavilan et al., 2014). Further, mental imagery vividness demonstrate that when consumers give automatic social attributions of
has been used as a stimulus in recent research on social commerce that personality to virtual agents, this may be sufficient to perceive human
uses the S–O-R framework (see Kim, 2015). In this paper, we view image others on a website or virtual environments. Research portrays social
quality as well as perceived mental imagery vividness as external presence as a key factor in how users relate to websites or electronic
stimulus (see Fig. 1: Conceptual framework) and hypothesize the platforms. Studies show that by boosting the feeling of being in the
following relationship between these variables. company of others, social presence positively affects trust in websites
(Cyr et al., 2009; Hollebeek et al., 2020), as well as decision-making
H1a. The level of image quality on social commerce positively impacts
processes within virtual communities and online purchase behaviors
perceived mental imagery vividness.
(Hess et al., 2009; Lu et al., 2016; Weisberg et al., 2011). Research
(Gefen and Straub, 2004; 2003; Karahanna and Straub, 1999; Kumar
2.1.3. Impact of reviews on social commerce
and Benbasat, 2002; Straub, 1994) indicates that social presence posi­
Prior research has shown that online reviews (Chevalier and May­
tively impacts purchase intention online. Hassanein and Head (2007)
zlin, 2006; Chintagunta et al., 2010) as well as textual advertisements
has explored the underlying mechanism and shown that higher levels of
affect consumers’ attitudes and decisions in e-commerce (Lee and Shin,
perceived social presence positively impact the perceived usefulness,
2014; Jiang et al., 2013; Limayem et al., 2000). In the social commerce
trust and enjoyment of shopping websites, leading to more favorable
context, customer reviews within online advertisements have been
consumer attitudes towards online shopping.
studied as a means of persuasion. They are employed to increase the
Previous research has associated mental imagery vividness with so­
attractiveness, conversion rate and sales volume in this context (Yuan
cial presence in different settings, like in TV audiovisual messages
et al., 2021). Further, customer reviews (textual and word-of-mouth
(Skalski and Tamborini, 2004) or online advertisements (Fortin and
from friends and family) that are not embedded in online advertise­
Dholakia, 2005). Further, the S–O-R framework has been used to
ments can create social influence that impacts user’s purchase behavior
examine the impact of mental imagery vividness as a stimulus input for
on social commerce platforms (Xu et al., 2019). Essentially, user prox­
an e-commerce (Eroglu et al., 2001) or social commerce website (Kim,
imity and interaction between users with strong ties allows customer
2015), in the analysis of the consumers’ response as shaped by the
reviews to leverage existing social relationships and influence purchases
experience of social presence. In line with this research, we consider
(Cao et al., 2021) leading to high conversion of users to buyers (Cao
social presence as the organisms’ (i.e., consumers’) internal state that
et al., 2020). Quality and valence of reviews also plays an important
explains the link between stimulus (mental imagery vividness) on the
role. High quality reviews with relevant, comprehensive, and accurate
response output (consumers’ purchase intention). Further, we take the
product-related information affects product evaluation and purchase
view that an organisms’ internal state can be composed of several var­
intention (Lee and Shin, 2014). Valence of reviews has been viewed
iables like cognitive and affective social presence that sequentially
using the S–O-R framework as a stimulus that has sequential effects on
impact each other (see conceptual framework in Fig. 1).
social interactions and trust as internal states, and purchase intention in
The distinction of affective and cognitive social presence was based
the social commerce context as the response (Li, 2019; Zhao et al.,
on previous work of Biocca et al. (2003) on the psychological component
2023). Recently, there has been a focus on studying linguistic features of
of social presence and its value for social media marketing (e.g., Hol­
reviews (Wang et al., 2023).
lebeek et al., 2020; Pechmann, 2015; Rundle-Thiele, 2015). Cognitive
In contrast to the extant literature, we focus on text review content in
social presence refers to an individual’s understanding capacity to
terms of whether they are experience-based or facts-based, while
confirm the meaning of other people’s interactions in social digital en­
keeping a consistent review valence. Experience-based reviews are
vironments” (Shen and Khalifa, 2008, p. 23). The authors explain that
associated with greater levels of mental imagery vividness than facts-
cognitive processes mediate the impact of immersion on the develop­
based reviews (Brewer, 2010). Similar to image quality, textual review
ment of presence extending mutual understanding. In the context of
context impacts mental imagery vividness since an individuals’ cogni­
social media, web interfaces impact social presence cognitively because
tion relies on both a verbal system based on speech and text as well as a
consumers aim to mitigate uncertainty in digital communications by
nonverbal system based on images (Paivio, 2008; 1991; Wang et al.,
resolving its ambiguities of product’s reviews in social media that are
2023). According to the dual coding theory, both images and text are
difficult to understand (DeMotta et al., 2016; Hu et al., 2017). Thus, we
inter-connected (Paivio, 2013), and might impact purchase intention by
expect mental imagery vividness in the social commerce context to
a similar mechanism (see Fig. 1: Conceptual framework). Hence, we
positively impact cognitive social presence such that consumers are
hypothesize that.
better able to understand opinions expressed in reviews.
H1b. Textual reviews with personal experiences, compared to reviews Further, affective social presence differs from cognitive social presence

6
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

in that it refers to the influence on an individual’s mood based on an

emotional connection aroused by interactions with other people in so­
cial digital environments (Shen and Khalifa, 2008). This differentiation
is important especially in social commerce context where reviews are
provided by social networks and are likely to be more influential than
reviews in e-commerce context. For example, reviews can play two clear
roles: a cognitive role about the level to which the reviews are under­
stood by the consumer and others in the network; and an affective role
about the level to which the reviews influence impressions and opinions
of consumers and others in the network. Thus, we can expect that if
reviews on social commerce are better understood then they will have a
greater influence on opinions of consumers and others; implying that
higher cognitive social presence in this context is likely to increase af­
fective social presence.
Further, academic literature has widely associated affective social
presence with changes in consumer’s attitudes and behaviors (e.g., Shen
and Khalifa, 2008; Skalski and Tamborini, 2004). For example, Skalski Fig. 2. Overview of the methodological framework.
and Tamborini (2004) manipulate mental imagery vividness by
exposing participants to a visual product using different screen sizes in a of items measuring latent constructs (15 items). In addition, we adopted
progressive continuum and show that stronger perceptions of affective the flexible cutoff values approach to evaluate the fit indexes of the SEM,
social presence are positively associated with larger screen sizes (high which takes into consideration the number of latent variables, sample
vividness), which in turn results in participants perceiving messages as size, among other parameters (Mai et al., 2021; Niemand and Mai, 2018)
more persuasive. as acknowledged in section 3.3. Analytical approach and testing of hy­
Applying the S–O-R framework to the above discussion, we view potheses. The primary data for Study 1 was collected in the United States
mental imagery vividness as an external stimulus, affective social pres­ through an online experimental survey using MTurk, which allows re­
ence and cognitive social presence as an organisms’ internal states, searchers to collect data at a low cost and in less time than traditional
purchase intention as the response output (see conceptual framework in survey methods (Hulland and Miller, 2018). Data collected with MTurk
Fig. 1) and hypothesize the following relationship between these is reliable and of similar quality to data collected with traditional
variables. methods (Park et al., 2020; Vazquez, 2021). Although collecting data
H2. The relation between mental imagery vividness and purchase intention with MTurk provides a large and diverse pool of participants with high
on social commerce is serially mediated by cognitive social presence and digital literacy, there is a high attrition rate of participants and lack of
affective social presence. participant’s attention (Aguinis et al., 2021; Kees et al., 2017). We fol­
lowed the recommendations of Aguinis et al. (2021) to retain valid re­
sponses from participants using MTurk in combination with Qualtrics to
3. Methodology
increase the validity of this data collection procedure following also
related studies (e.g., Toteva et al., 2021). First, we filtered out inatten­
We followed the three steps to carry out both studies testing the
tive participants who did not pass the attention check (39 participants
research hypotheses. First, we designed two stimuli for simple experi­
removed). Second, we filtered out responses where no input activity was
mental comparisons between groups. Second, we determined the mini­
recorded (i.e., clicks or keyboard hit) while responding to the ques­
mum sample size for each treatment group and collected the data by
tionnaire (30 participants removed) as these responses were suspect of
randomly presenting one of the two stimuli to each participant. Before
being generated from web-robots, which should be avoided according to
proceeding to the analysis of data, we carried out data validation rec­
Aguinis et al. (2021). Third, we filtered out incomplete questionnaires
ommendations according to the data collection tool. Third, we ran an­
alyses to confirm reliability and discriminant validity of measures, as
well as verified that the random assignment resulted in equivalent Table 2
treatment groups. We test the research hypotheses using regression and Descriptive statistics of the sample and correlation matrix.
path analyses, alongside a structural equation model (SEM). An over­ M SD PI A I E VIV ASP
view of the methodological framework is described in Fig. 2. PI 4.1 1.5 –
A 35.9 10.5 .14a –
I 4 .00 .09 –
3.1. Study 1 E 13 .14a .12′ .15a –
VIV 4.9 1.4 .51*** .19** .00 .07 –
Study 1 explores the effects of image quality on perceived mental ASP 4.7 1.2 .33*** .07 − .03 .05 .15a –
imagery vividness (H1a) and their effect on subsequent perceptions of CSP 5.0 1.0 .25*** .06 − .02 − .04 .22** .64***

cognitive social presence and affective social presence on purchase PI: Scale of purchase intention, A: Age (years), I: Income level (9 ordinal cate­
intention (H2). goriesb), E: Education level (15 ordinal categoriesc), VIV: Scale of mental im­
agery vividness. Significance level: ***p < .001; **p < .01; ap < .05; ‘p < .10.
a
3.1.1. Sample and measurements Legend: M = Mean and SD=Std. Deviation, for the categorical variables
To determine the minimum number of observations in our experi­ Income level and Education level M = Mode.
b
ment, we followed Kline (2015) who recommends collecting at least 100 Less than $15,000 USDs (1), $15,000 - $24,999 USDs (2), $25,000 - $34,999
USDs (3), $35,000 - $49,999 USDs (4), $50,000 - $74,999 USDs (5), $75,000 -
observations for each treatment group. Our final sample of 214 partic­
$99,999 USDs (6), $100,000 - $149,999 USDs (7), $150,000 - $199,999 USDs
ipants meets this criterion. Although there is no consensus about a
(8), and $200,000 USDs or more (9).
unique method to calculate the sample size for structural equation c
None (1), 1st - 4th grade (2), 5th - 6th grade (3), 7th - 8th grade (4), 9th grade
modeling (SEM), the analytical approach used for testing of hypotheses (5), 10th grade (6), 11th grade (7), High school graduate (8), Some college, no
(see section 3.3); we verified the sample size meets the guidelines from degree (9), Associate degree, occupational (10), Associate degree, academic
Bentler and Chou (1987). That is, at least 75 observations per treatment (11), Professional degree (12), Bachelor’s (13), Master’s (14), Doctoral degree
group following their recommendation of multiplying by 5 the number (15).

7
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

(20 participants removed). Finally, we retained a final sample of 214 Table 3

women participants (MAge = 35.9 years old, SD = 10.5), representing Confirmatory factor analysis and items for each latent construct.
70.6% of the raw data sample (303 participants) which is within the Measures Factor AVE CR Cronbach
range of missing data points due to inattentive participants, web-robots, load α
and high attrition rate reported by Aguinis et al. (2021 p. 826). The Purchase intention .77 .93 .93
majority of participants had an age between 25 and 54 years old (85% of
The likelihood of me purchasing this .826
the sample), which is in line with the age range of largest group of online Tweed Jacket isa
shoppers in USA (52.9%) obtained by eMarketer (2020). Our sample If I were going to purchase an apparel .903
underrepresents online shoppers below 24 years old and above 55 years product, I would consider buying
old, as well as it overrepresents women aged between 25 and 44 years Beautiful Social Apparelb
If I were shopping for an apparel brand, .937
old; however, the shape of the age distribution histogram of our sample
the likelihood I would purchase
is in line with that obtained by eMarketer (an age distribution com­ Beautiful Social Apparel is highb
parison is in the appendix). We focus on middle-aged women because of My willingness to buy this Tweed Jacket .877
their great power and influence on consumer’s decisions (Silverstein and from Beautiful Social Apparel isa
Sayre, 2009). As caregivers for children and the elderly in most societies, Affective social presence .55 .78 .77
women buy on behalf of the people who live in their households, and
I am influenced by the reviews of othersb .764
they influence purchasing decisions in their social circles affecting 75% The reviews of others affect my .799
of consumer purchases (Berg and Amed, 2020; Brennan, 2015; Statista, impressions about apparel productsb
2019). I think people’s reviews of apparel .641
products in this page affect each other’s
After being exposed to a mocked-up social media page promoting a
impressions about apparel productsb
Tweed jacket, participants reported their purchase intentions in a scale
originally developed by Dodds et al. (1991), also used by Lee and Cognitive social presence .61 .82 .81

Watkins (2016) to measure purchase intentions of luxury brands. Next, People in this page understand each otherb .640
we measured the two dimensions of social presence using adapted items I understand the others’ opinions on the .818
apparel productsb
originally developed by Biocca et al. (2001) to measure cognitive and
The other individuals’ thoughts on the .848
affective social presence, as proposed by Shen and Khalifa (2008). apparel products are clear to meb
Perceived vividness was measured with an adapted scale from Bone and
Mental imagery vividness .62 .89 .89
Ellen (1992), which has also been used in related studies to measure
mental imagery vividness (e.g., Babin and Burns, 1998; Lee, 2012). Unclear: Clear .724
Pale: Vivid .823
Table 2 offers a wider description of the sample and correlations be­
Mild: Intense .755
tween variables. Blunt: Sharp .773
In line with established procedures to ensure the validity and reli­ Undefined: Well-defined .839
ability of measures, we conducted a confirmatory factor analysis (CFA)
SRMR = 0.058; CFI = 0.937; TLI = 0.921; RMSEA = 0.084; χ2 = 212.2; p < .001;
on our mediating and dependent variables. Results were above the df = 84; χ2/df = 2.53.
minimum cutoff values of 0.5 for each average variance extracted, 0.6 All items measured with 7-point scales.
for each factor loading, and 0.7 for each composite reliability score a
(very low – very high).
(Fornell and Larcker, 1981; Voorhees et al., 2016). We verified that the b
(strongly disagree – strongly agree).
highest squared correlation between latent constructs was smaller than
the minimum average variance extracted (ρ2 = 0.41 < AVEmin = 0.55), Sightengine software to obtain the image’s sharpness index (Boutilier
thus confirming discriminant validity (Rönkkö and Cho, 2022). Table 3 et al., 2019, p. 62) and the function of “image attributes” from Google
shows a summary of the results and fit indexes of the CFA model, which Cloud Vision API and computed a colorfulness index based on Li and Xie
comply with the flexible cutoff values suggested by Niemand and Mai (2020, p. 5). A comparison based on these indexes in Table 3 shows that
(2018). the manipulated image resulted with 22.9% less colorfulness and 10.2%
less sharpness than the original image. The experimental treatment
3.1.2. Manipulations and procedure consisted in randomly presenting one of these two image sets (see Fig. 3)
To prepare the stimuli, for each of the two treatment groups we to each participant. Table 4 shows the characteristics of image
created a set of images depicting two versions of a mocked-up social manipulation.
commerce site on Instagram selling a Tweed jacket for women. We A simple experimental design with two levels of image quality (high
emulated a mobile interface of Instagram because it is one of the most vs low), was used to test the research hypotheses. The experimental
popular social commerce platforms in which we could emulate a post of survey has four sections: in the first section, participants visit a social
model wearing an apparel product, textual comments from other social commerce site that promotes a Tweed jacket, worn by a model, for at
media users, and the checkout process (Herzallah et al., 2022; Saprikis least 20 s before the ‘next’ button appears to advance to the second
and Avlogiaris, 2021; Yahia et al., 2018). Because apparel is one of the section. This time threshold aims to set a lower boundary to the par­
main social commerce categories sold through Instagram (Jin and Ryu, ticipant’s exposure time in line with the work of Vazquez (2020). The
2020), we selected a Tweed jacket as we believe is characteristic of fe­ second section included the scales comprising the dependent variable of
male online shoppers, our targeted sample (Berg and Amed, 2020; Sil­ purchase intention, whereas the third section included the scales cor­
verstein and Sayre, 2009; eMarketer, 2020). To manipulate image responding to the mediating variables of mental imagery vividness and
quality, we used Photoshop software to decrease the image set sharpness the two social presence dimensions proposed in this study (cognitive and
(Starke et al., 2021) and colorfulness (Labrecque, 2020). We used the

8
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

Table 5
Descriptive statistics of the treatment groups.
Treatment group I II

Sample size 110 104

Image quality Low High

Age 34.1 (9.5) 37.9 (11.1)

Income 4 4
Education 13 13
Purchase intention 4.0 (1.5) 4.3 (1.5)
Affective social presence 4.7 (1.1) 4.7 (1.3)
Cognitive social presence 5.0 (1.0) 5.0 (1.0)
Mental imagery vividness 4.5 (1.5) 5.2 (1.2)

Legend: Mean (SD), for categorical variables only the mode is provided.

3.1.3. Analytical approach and testing of hypotheses

To test the research hypotheses, we used a structural equation model
(SEM) using mean composite scores of the items pertaining to each
latent construct of mental imagery vividness, cognitive social presence,
affective social presence, and purchase intention. The model evaluates
the effects of mental imagery vividness induced by image quality on
purchase intention through a serial mediation process involving cogni­
tive and affective social presence. We assessed the statistical significance
of the model with 10,000 bootstrap samples to create bias-corrected
confidence intervals at 95%, with heteroscedasticity-consistent stan­
dard errors. The manipulation was coded as a dummy variable following
the recommendation of Hayes (2018). The fit indexes of the SEM model
for testing the research hypotheses (χ2 = 1.76, p = .62, RMSEA = 0.018,
SRMR = 0.018) were within the recommend cutoff values recommended
by Niemand and Mai (2018), which indicates a good model fit. Fig. 4
represents the full model used to test the research hypotheses.
According to the test, greater image quality leads to a higher mental
imagery vividness (effect = 0.69, p < .001), which indicates support for
H1a. Next, mental imagery vividness leads to higher cognitive social
presence (effect = 0.16, p = .008), which in turn significantly predicts
affective social presence (effect = 0.75, p < .001). Consequently, af­
fective social presence affects purchase intention (effect = 0.33, p =
.004). The relation between mental imagery vividness and purchase
intention on social commerce is serially mediated by cognitive social
presence followed by affective social presence (indirect effect = 0.04, p
Fig. 3. Image manipulation for Study 1
= .039, 95% CI [0.014, 0.093]), which supports H2. The total effect
Notes: The mockups used for both studies are consistent with previous empir­
results significant and positive indicating that the experimental
ical studies (see Kanani and Glavee-Geo (2021, p. 15). The purpose of Study 1 is
manipulation of image quality produces a meaningful effect on purchase
to evoke imagery vividness based on image quality. For Study 2 only the
low-quality image (Fig. 3a) was used, the comments section was manipulated to intention. Table 6 shows the results of regressions and Table 7 presents
evoke imagery vividness based on review type (facts vs personal experience). path analysis results.

Table 4 3.2. Study 2

Image manipulation.
Treatment group Colorfulness Sharpness Study 2 explores the effects of text content in product reviews on
index index perceived mental imagery vividness (H1b) and their effect on
I. Low image quality (manipulated image Monotonic .37 Blurry .88
set)
II. High image quality (original image set) Colorful .48 Sharp .98
Difference .11 (22.9%) .10 (10.2%

affective). Because we analyze the phenomenon from the

Stimulus-Organism-Response paradigm, which separates the stimulus
from the organism’s internal state, we consider that a manipulation
check is not necessary as the relationship between the Stimulus and the
organism’s internal state are within our model (Hauser et al., 2018). In
the fourth section participants shared their demographic information,
including age, level of education, and income. Although the random Fig. 4. Impact of Image Quality through Mental Imagery Vividness on Purchase
assignment of treatment balanced most demographic variables, we Intention: Mediating Role of Cognitive and Affective Social Presence
included these variables as covariates in the analyses. Fig. 3 shows the Significance level: ***p < .001; **p < .01; *p < .05; ‘p < .10SRMR = 0.018; CFI
stimulus of each treatment group and Table 5 shows descriptive statis­ = 1.000; TLI = 1.042; RMSEA = 0.000; χ 2 = 1.76; p = .62; df = 3; χ2/df
tics of each treatment group. = 0.586.

9
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

Table 6
Regressions and covariates Study 1.
Antecedent Consequent

VIV(M1) CSP(M2) ASP(M3) PurchInt(Y)

Coef SE t p Coef SE t p Coef SE t p Coef SE t p

IQ(X) .69 .18 3.71 <.001

VIV(M1) .16 .05 3.16 <.01 − .00 .05 − .07 .95 .51 .07 7.81 <.001
CSP(M1) .75 .06 11.70 <.001 − .01 .11 − .12 .91
ASP(M2) .33 .10 3.48 <.001
Age .02 .01 2.26 .03 .00 .01 .43 .67 .00 .00 .48 .63 .00 .01 .35 .72
Income − .03 .04 − .66 .50 − .01 .03 − .17 .87 − .02 .03 − .65 .51 − .00 .04 − .18 .85
Edu .05 .04 1.16 .25 − .02 .03 − .90 .37 .04 .03 1.47 .14 .06 .04 1.59 .11
Constant 3.44 .54 6.32 <.001 4.47 .45 10.0 <.001 .49 .50 .99 .32 − .65 .69 − .93 .35

Observations 214

R2 .10 .05 .41 .33

Adjusted R2 .08 .03 .39 .31
Residual SE 1.32 (df = 209) .99 (df = 209) .92 (df = 208) 1.26 (df = 207)
F statistic F(4,209) = 5.82, p < .001 F(4,209) = 2.88, p = .023 F(5,208) = 29.29, p < .001 F(6,207) = 17.29, p < .001

PI: Scale of purchase intention, ASP: Affective social presence, CSP: Cognitive social presence, IQ: Image quality treatment (dummy variable), Covariates: Age (years),
Income level, and Education Level.

participants removed), and incomplete questionnaires (23 participants

Table 7 removed). Hence, we retained a final sample of 214 women participants
Results of path analysis Study 1.
(MAge = 34.2 years old, SD = 9.9) in Study 2, representing 69.3% of the
Description of the effect (path) coeff se p- LLCI ULCI raw data sample (309 participants) in line with the percentages of
value
missing data points due to inattentive participants, web-robots, and high
IQ → VIV → PI .352 .108 .001 .184 .620 attrition rate (Aguinis et al., 2021 p. 826). The measurement scales used
IQ → VIV → CSP → PI − .001 .017 .933 − .041 .028
IQ → VIV → ASP → PI − .001 .013 .955 − .032 .020
IQ → CSP → ASP → PI .170 .084 .043 .041 .392
IQ → VIV → CSP → ASP → PI .027 .015 .077 .011 .082 Table 9
Total effect 1.058 .192 .000 .648 1.392 Confirmatory factor analysis and items for each latent construct.
VIV → CSP → ASP → PI (path tested .040 .019 .039 .014 .093 Measures Factor AVE CR Cronbach
for H2) load α
LLCI = lower limit confidence interval; ULCI = upper limit confidence interval. Purchase intention .79 .94 .94
Based on 10,000 bootstrap samples.
The likelihood of me purchasing this .834
Tweed Jacket isa
subsequent perceptions of cognitive social presence and affective social If I were going to purchase an apparel .919
presence on purchase intention (H2). product, I would consider buying
Beautiful Social Apparelb
If I were shopping for an apparel brand, .930
3.2.1. Sample and measurements the likelihood I would purchase
Following the procedure of Study 1, we filtered out inattentive par­ Beautiful Social Apparel is highb
ticipants (34 participants removed), responses from web-robots (32 My willingness to buy this Tweed Jacket .884
from Beautiful Social Apparel isa

Affective social presence .53 .76 .75

Table 8
Descriptive statistics of the sample and correlation matrix. I am influenced by the reviews of others b
.782
The reviews of others affect my .761
M SD PI A I E VIV ASP
impressions about apparel productsb
PI 4.1 1.6 – I think people’s reviews of apparel .602
A 34.2 9.9 − .01 – products in this page affect each other’s
I 5 .00 .15a – impressions about apparel productsb
E 13 .08 .03 .24*** –
Cognitive social presence .59 .81 .80
VIV 4.3 1.6 .54*** .05 .04 .13′ –
ASP 4.7 1.1 .35*** .02 .03 .07 .21** – People in this page understand each otherb .652
CSP 5.0 1.0 .27*** .11′ − .02 − .07 .19** .58*** I understand the others’ opinions on the .811
apparel productsb
PI: Scale of purchase intention, A: Age (years), I: Income level (9 ordinal cate­ The other individuals’ thoughts on the .819
goriesb), E: Education level (15 ordinal categoriesc), VIV: Scale of mental im­ apparel products are clear to meb
agery vividness. Significance level: ***p < .001; **p < .01; ap < .05; ‘p < .10.
a Mental imagery vividness .67 .91 .90
Legend: M = Mean and SD=Std. Deviation, for the categorical variables
Income level and Education level M = Mode. Unclear: Clear .795
b
Less than $15,000 USDs (1), $15,000 - $24,999 USDs (2), $25,000 - $34,999 Pale: Vivid .848
USDs (3), $35,000 - $49,999 USDs (4), $50,000 - $74,999 USDs (5), $75,000 - Mild: Intense .778
Blunt: Sharp .751
$99,999 USDs (6), $100,000 - $149,999 USDs (7), $150,000 - $199,999 USDs
Undefined: Well-defined .876
(8), and $200,000 USDs or more (9).
c
None (1), 1st - 4th grade (2), 5th - 6th grade (3), 7th - 8th grade (4), 9th grade SRMR = 0.064; CFI = 0.938; TLI = 0.923; RMSEA = 0.085; χ2 = 213.6; p < .001;
(5), 10th grade (6), 11th grade (7), High school graduate (8), Some college, no df = 84; χ2/df = 2.54.
degree (9), Associate degree, occupational (10), Associate degree, academic All items measured with 7-point scales.
a
(11), Professional degree (12), Bachelor’s (13), Master’s (14), Doctoral degree (very low – very high).
b
(15). (strongly disagree – strongly agree).

10
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

in Study 2 were the same used ones used in Study 1. Table 8 describes Table 11
the sample and correlations. Descriptive statistics of the treatment groups.
To ensure the validity and reliability of measures, we conducted a Treatment group I II
CFA following Study 1. Results were above the minimum cutoff values of
Sample size 110 104
0.5 for each average variance extracted, 0.6 for each factor loading, and
0.7 for each composite reliability score and the highest squared corre­ Text content review Facts Personal experience

lation between latent constructs was smaller than the minimum average Age 34.4 (10.4) 34.1 (9.5)
variance extracted (ρ2 = 0.33 < AVEmin = 0.53), confirming discrimi­ Income 4 5
Education 13 13
nant validity in Study 2. Table 9 shows a summary of the results and fit
Purchase intention 4.2 (1.7) 4.0 (1.5)
indexes of the CFA model, which comply with the flexible cutoff values Affective social presence 4.7 (1.2) 4.7 (1.1)
(Niemand and Mai, 2018). Cognitive social presence 4.9 (1.1) 5.0 (1.0)
Mental imagery vividness 4.1 (1.6) 4.5 (1.5)
3.2.2. Manipulations and procedure Legend: Mean (SD), for categorical variables only the mode is provided.
Because the purpose of Study 2 is to induce mental imagery vividness
through the manipulation of textual comments, we used the image set
with low quality from Study 1 as a basis. This makes participants focus
on the textual comments of other social media users rather than on the
image of the woman wearing an apparel product. Hence, we created two
treatment groups by manipulating only the text content of reviews.
Based on the work of Itzchakov et al. (2020), the treatment group I had
text reviews framed as facts and the treatment group II had text reviews
framed as personal experiences. Table 10 shows the manipulation of text
content presented as comments from other social media users and the
experimental treatment consisted in randomly presenting one of these
two variations of textual comments to each participant.
A simple experimental design with two text content frames (facts vs Fig. 5. Impact of Review Type through Mental Imagery Vividness on Purchase
personal experience), served to test the research hypotheses in Study 2. Intention: Mediating Role of Cognitive and Affective Social Presence
The experimental survey followed the procedure and the four sections Significance level: ***p < .001; **p < .01; *p < .05; ‘p < .10SRMR = 0.019; CFI
described in Study 1. However, the random assignment of treatment = 0.993; TLI = 0.947; RMSEA = 0.046; χ 2 = 4.37; p = .22; df = 3; χ2/df
balanced in Study 2 all demographic variables between groups (p > .10), = 1.456.
thus no covariate was included the analysis. Table 11 shows descriptive
statistics of each treatment group. indicates a good model fit. Fig. 5 represents the full model used to test
the research hypotheses.
3.2.3. Analytical approach and testing of hypotheses Text content framed as a personal experience leads to a higher
To test the research hypotheses, we followed the SEM analytical mental imagery vividness (effect = 0.39, p = .06) which indicates sup­
approach used in Study 1. The model evaluates the effects of mental port for H1b at p < .10 level, it is also important to consider that our data
imagery vividness induced by text content (e.g., dummy variable 0 = collection sample could reduce effect sizes estimates by up to 40%
facts vs 1 = personal experience) on purchase intention through a serial (Aguinis et al., 2021). Mental imagery vividness leads to higher cogni­
mediation process involving cognitive and affective social presence. The tive social presence (effect = 0.13, p = .004), which strongly impacts
fit indexes of the SEM model for testing the research hypotheses (χ2 = affective social presence (effect = 0.63, p < .001). Consequently, af­
4.37; p = .22; RMSEA = 0.046; SRMR = 0.019) were within the range of fective social presence affects purchase intention (effect = 0.32, p =
cutoff values recommended by Niemand and Mai (2018), which .001). The relation between mental imagery vividness and purchase
intention on social commerce is serially mediated by cognitive social
presence followed by affective social presence (indirect effect = 0.03, p
Table 10 = .032, 95% CI [0.009, 0.062]), which supports H2. The total effect
Text manipulation for Study 2. results significant and positive indicating that the experimental
Product’s Treatment Group manipulation of text content as a product review produces a meaningful
review effect on purchase intention. Table 12 shows the results of regressions
I. Facts II. Personal Experience
and Table 13 presents path analysis results.
1 This Tweed Jacket has some Have wanted one for a while.
upgraded properties, but it has This Tweed Jacket has some
4. Discussion
all the properties an ordinary upgraded properties, but is more
user needs. than adequate for what I need
2 This Tweed Jacket can be a great My girlfriend LOVES her Tweed Images and customer reviews are widely accepted as key de­
gift. You can wear it under the Jacket. When I told her she could terminants of consumer behavior in the context of social media and e-
rain and enjoy all the features, wear it under the rain, she was commerce, however their role in social commerce has been under-
including getting stuff inside its amazed. She loves all the
multiple secret pockets. features, including getting all
studied. In this research, we expand existing knowledge on the impact
her stuff inside its multiple of images and customer reviews on consumer decision-making by
secret pockets examining the important role of image quality and textual review type
3 It is important to do comparative I have had this Tweed Jacket for (personal experiences versus facts) in enhancing purchase intention on
shopping between this Tweed 5 days. I’m glad I did my
social commerce platforms. Specifically, we contribute to the current
Jacket and others, although it comparative shopping, although
takes a lot of time. This is a very it took a lot of time. This is the literature by linking the influence of image quality and review type with
good Tweet Jacket. best Tweet Jacket I have ever mental imagery vividness and social presence. We provide useful in­
had. sights to managers by demonstrating how to enrich mental imagery
Adapted from (Itzchakov et al., 2020), supplementary material p. 15 and p. 23 vividness to impact purchase intention as a socially influenced decision.
valence-consistent vividness condition (personal experience) and p. 14 and p. 25
valence-consistent facts condition.

11
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

Table 12
Regressions and covariates Study 2.
Antecedent Consequent

VIV(M1) CSP(M2) ASP(M3) PurchInt(Y)

Coef SE t p Coef SE t p Coef SE t p Coef SE t p

TC(X) .39 .18 2.07 .06

VIV(M1) .13 .04 2.93 <.01 .07 .04 1.65 .09 .50 .06 8.3 <.001
CSP(M1) .63 .06 9.98 <.001 .08 .11 .73 .47
ASP(M2) .32 .10 3.17 <.01
Age .01 .01 .74 .45 .01 .01 1.62 .10 − .01 .01 − .95 .34 − .01 .01 − .77 .44
Income .00 .05 − .00 .99 − .01 .04 − .32 .74 .01 .03 .41 .68 − .02 .05 − .37 .71
Edu .08 .05 1.64 .09 − .04 .03 − 1.31 .19 .05 .03 1.62 .11 .01 .04 .23 .82
Constant 2.92 .64 4.53 <.001 4.53 .44 10.3 <.001 .95 .49 1.96 .05 .25 .71 .36 .72

Observations 214

R2 .03 .06 .36 .35

Adjusted R2 .02 .04 .34 .33
Residual SE 1.54 (df = 209) 1.01 (df = 209) .91 (df = 208) 1.31 (df = 207)
F statistic F(4,209) = 1.84, p = .12 F(4,209) = 3.18, p = .01 F(5,208) = 23.05, p < .001 F(6,207) = 18.86, p < .001

PI: Scale of purchase intention, ASP: Affective social presence, CSP: Cognitive social presence, TC: Text content review (dummy variable), Covariates: Age (years),
Income level, and Education Level.

opinions in consumers (affective aspect). Thus, the sequential role of

Table 13 cognitive and affective social presence is crucial for understanding the
Results of path analysis Study 2.
role of social influence in the social commerce context.
Description of the effect (path) coeff se p- LLCI ULCI In this study we provide a shift of focus in the traditional S–O-R
value
paradigm, which mainly emphasizes the experimental examination of
TC → VIV →PI .195 .098 .046 − .004 .403 imagery elaboration in related studies, as well as the focus on one single
TC → VIV → CSP → PI .004 .008 .603 − .005 .031 dimension of social presence as a mediator of mental imagery vividness.
TC → VIV → ASP → PI .008 .008 .302 − .001 .037
TC → CSP → ASP → PI .077 .049 .117 .007 .228
Our study offers empirical evidence of the effect of two dimensions of
TC → VIV → CSP → ASP → PI .010 .007 .159 .002 .037 social presence on the mental imagery vividness—purchase intention
Total effect .794 .166 .000 .474 1.144 relationship. The results confirm that people are more attracted to
VIV → CSP → ASP → PI (path tested .026 .012 .032 .009 .062 apparel products presented vividly in social media sites by arousing the
for H2)
perceptions of cognitive and affective social presence.
LLCI = lower limit confidence interval; ULCI = upper limit confidence interval.
Based on 10,000 bootstrap samples.
4.2. Managerial implications
4.1. Theoretical implications
The findings of our paper provide valuable practical marketing
Following the S–O-R paradigm, we develop a conceptual framework suggestions to managers of social commerce pages and platforms in
where-in image quality and review type on social commerce evoke terms of managing image quality and customer reviews. The findings
mental imagery vividness, which in turn influences consumers’ cogni­ indicate that managers should try to obtain textual reviews with per­
tive, affective, and behavioral responses. Furthermore, in this paper we sonal experiences from customers. Managers need to implement mar­
explore the mediating role of cognitive and affective social presence on keting strategies to ensure that customers are encouraged to post
how mental imagery vividness influences consumers’ purchase inten­ reviews with personal experiences. For example, it might be useful to
tion. Our findings offer empirical support for the induction of mental send a post-purchase survey asking customers to describe their experi­
imagery vividness by images and reviews, as well as the role of mental ence with the product purchased and post this reply as a textual review
imagery vividness in shaping subsequent consumer response. We show on the social commerce site. Alternatively, a social commerce platform
that image quality and review type share a similar mediation mechanism could as a policy mention that they want customers to provide reviews
since both impact the level of mental imagery vividness experienced on that are experience-based as opposed to being based on a factual list of
social commerce platforms. product attributes.
Consistent with prior research on mental imagery vividness in con­ Similarly, managers need to implement marketing strategies to
texts like online advertisements (Fortin and Dholakia, 2005), and TV encourage customers to upload high quality product images after pur­
audiovisual messages (Skalski and Tamborini, 2004), we show that chase. Managers of social media and social commerce platforms need to
imagery vividness experienced by users on social commerce platforms ensure that it is technically possible for users to upload high quality
leads to a feeling of social presence. Whereas social presence evoked by images. For instance, Instagram users usually complain that their
mental imagery is an important driver of purchase intention in the uploaded photos are downgraded into low quality formats by the plat­
e-commerce context (Hess et al., 2009; Lu et al., 2016; Weisberg et al., form. This happens because Instagram limits the space provided to each
2011), we disclose a path-dependent mechanism by which mental im­ image, automatically resizing, and compressing images to the maximum
agery vividness impacts purchase intention in the social commerce setting allowed. Similar issues are common in other social media plat­
context. Specifically, we reveal the sequential role of cognitive and af­ forms like Twitter and Facebook. Therefore, technical solutions to
fective social presence as mediators of the relation between mental overcome limitations of space need yet to be advanced to ensure a high
imagery vividness and purchase intention. This is particularly important quality of product images. Alternatively, defining an internal parame­
in social commerce content because we conceptualize and measure so­ trization policy for images may also mitigate the effects of image
cial presence in terms of the level to which reviews are understood by compression in these platforms.
consumers (cognitive aspect), and the level to which they influence Further, our findings provide practical implications for managers in
understanding the underlying path by which mental imagery vividness

12
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

influences purchase intention in the social commerce context. These consultancies, stock ownership, or other equity interest; and expert
findings highlight that when users experience higher mental imagery testimony or patent-licensing arrangements), or non-financial interest
vividness, they are likely to understand positive reviews better and this (such as personal or professional relationships, affiliations, knowledge
eventually influences the users’ opinion about the product, which in or beliefs) in the subject matter or materials discussed in this
turn provides more closure in terms of purchase intention. This is a manuscript.
crucial role of mental imagery vividness, which is particularly important
in the social commerce context where reviews become the core of social Data availability
influence. This implies that managers are likely to get more advantage
from positive reviews when product images are of high quality and re­ Data will be made available on request.
views are based on personal experiences. So, the conversion rate of users
to buyers would be higher for social commerce platforms where positive Acknowledgements
reviews are complemented by high quality images and experience-based
reviews. The authors are grateful for the valuable feedback received from the
professors Dennis Herhausen, Charles Spence, and Reto Felix in earlier
5. Limitations and future research versions of the paper.

We note some limitations in this research. First, our results are based References
on self-reported measures. Second, we used only one social media
platform (Instagram), as fan pages on this platform are popular for e- Aguinis, H., Villamor, I., Ramani, R.S., 2021. MTurk research: review and
recommendations. J. Manag. https://doi.org/10.1177/0149206320969787.
retailing. Future studies might replicate this research in a wider selec­ Andrade, J., May, J., Deeprose, C., Baugh, S.J., Ganis, G., 2014. Assessing vividness of
tion of social media sites and use indirect measures instead of self- mental imagery: the plymouth sensory imagery questionnaire. Br. J. Psychol. 105
reported measures. Moreover, as this study focuses only on one type https://doi.org/10.1111/bjop.12050.
Babin, L.A., Burns, A.C., 1998. A modified scale for the measurement of communication-
of apparel product (a dress), other experimental interventions might evoked mental imagery. Psychol. Market. https://doi.org/10.1002/(SICI)1520-6793
want to introduce a larger array of products to check for potential dif­ (199805)15:3<261::AID-MAR4>3.0.CO;2-8.
ferences brought by each type of product. Third, as mentioned previ­ Bentler, P.M., Chou, C.P., 1987. Practical issues in structural modeling. Socio. Methods
Res. 16 https://doi.org/10.1177/0049124187016001004.
ously, this study induced participants to examine the manipulated Berg, A., Amed, I., 2020. The State of Fashion 2020. McKinsey&Company.
websites for a minimum of 20 s, which may have altered participants’ Biocca, F., Harms, C., Burgoon, J.K., 2003. Toward a more robust theory and measure of
natural behavior considering that their attention span in social media social presence: review and suggested criteria. Presence Teleoperators Virtual
Environ. 12, 456–480. https://doi.org/10.1162/105474603322761270.
may be much shorter. Additionally, the manipulated websites used in
Biocca, F., Harms, C., Gregg, J., 2001. The networked minds measure of social presence:
this study are static and do not include dynamic customer-to-customer pilot test of the factor structure and concurrent validity. In: 4th Annual International
interactions which are environmental characteristics of social com­ Workshop on Presence.
merce (Liu et al., 2021). Further research could test whether the results Bitner, M.J., 1992. Servicescapes: the impact of physical surroundings on customers and
employees. J. Market. 56 https://doi.org/10.1177/002224299205600205.
of this research change in a dynamic environment. Fourth, because we Bone, P.F., Ellen, P.S., 1992. The generation and consequences of communication-evoked
recruited participants using MTurk, the sample in this study is conse­ imagery. J. Consum. Res. 19, 93–104. https://doi.org/10.1086/209289.
quently limited to a segment of the population with high digital literacy. Boutilier, R.M., Ahn, Y.J., Park, J.S., Lee, H., 2019. Development of an auto-
interchangeable multi-pinhole array for confocal laser scanning microscopy systems
Future studies could consider people with a wider range of demographic using precision laser processing. Opt Laser. Technol. 116, 58–67. https://doi.org/
characteristics to shed further light on the internal motivations for retail 10.1016/j.optlastec.2019.03.004.
purchase intentions in a physical setting. Brennan, B., 2015. Top 10 Things Everyone Should Know about Women Consumers.
Forbes Online.
Further research could use other methods (e.g., natural field exper­ Brewer, W.F., 2010. Memory for randomly sampled autobiographical events. In:
iments, secondary data sources) without setting a minimum time for Remembering Reconsidered. https://doi.org/10.1017/cbo9780511664014.004.
interacting with the stimuli, to provide more realism. Future studies may Busalim, A.H., Hussin, A.R.C., 2016. Understanding social commerce: a systematic
literature review and directions for further research. Int. J. Inf. Manag. https://doi.
also explore how to evoke mental imagery vividness by the manipula­ org/10.1016/j.ijinfomgt.2016.06.005.
tion of sounds or environmental music as this study was limited to Cao, H., Chen, Z., Cheng, M., Zhao, S., Wang, T., Li, Y., 2021. You recommend, I buy.
exploring text and images. As the present study only considered the Proc ACM Hum. Comput. Interact. 5 https://doi.org/10.1145/3449141.
Cao, H., Chen, Z., Xu, F., Wang, T., Xu, Y., Zhang, L., Li, Y., 2020. When your friends
effects of social presence in the form of positive customers’ appraisals of
become sellers: an empirical study of social commerce site Beidian. In: Proceedings
one product, further replies might focus on the effect of social presence of the 14th International AAAI Conference on Web and Social Media. ICWSM.
when the customers’ reviews, ratings, or opinions are negative or mixed. https://doi.org/10.1609/icwsm.v14i1.7281, 2020.
Chau, P.Y.K., Au, G., Tam, K.Y., 2000. Impact of information presentation modes on
online shopping: an empirical evaluation of a broadband interactive shopping
Funding service. J. Organ. Comput. Electron. Commer. 10.
Chevalier, J.A., Mayzlin, D., 2006. The effect of word of mouth on sales: online book
No. reviews. J. Market. Res. https://doi.org/10.1509/jmkr.43.3.345.
Ching, R.K.H., Tong, P., Chen, J.S., Chen, H.Y., 2013. Narrative online advertising:
identification and its effects on attitude toward a product. Internet Res. 23 https://
Availability of stimulus material doi.org/10.1108/IntR-04-2012-0077.
Chintagunta, P.K., Gopinath, S., Venkataraman, S., 2010. The effects of online user
reviews on movie box office performance: accounting for sequential rollout and
Yes. aggregation across local markets. Market. Sci. 29, 944–957. https://doi.org/
10.1287/mksc.1100.0572.
Data code availability Cox, T., Park, J.H., 2014. Facebook marketing in contemporary orthodontic practice: a
consumer report. J. World Fed. Orthod. 3 https://doi.org/10.1016/j.
ejwf.2014.02.003.
Upon request. Coyle, J.R., Thorson, E., 2001. The effects of progressive levels of interactivity and
vividness in web marketing sites. J. Advert. 30, 65–77. https://doi.org/10.1080/
00913367.2001.10673646.
Declaration of competing interest Cyr, D., Head, M., Larios, H., Pan, B., 2009. Exploring human images in website design: a
multi-method approach. MIS Q. 33, 539–566 (Article).
The authors whose names are listed immediately below certify that De Vries, L., Gensler, S., Leeflang, P.S.H., 2012. Popularity of brand posts on brand fan
pages: an investigation of the effects of social media marketing. J. Interact. Market.
they have NO affiliations with or involvement in any organization or
26, 83–91. https://doi.org/10.1016/j.intmar.2012.01.003.
entity with any financial interest (such as honoraria; educational grants;
participation in speakers’ bureaus; membership, employment,

13
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

DeMotta, Y., Chao, M.C. ho, Kramer, T., 2016. The effect of dialectical thinking on the Karahanna, E., Straub, D.W., 1999. The psychological origins of perceived usefulness and
integration of contradictory information. J. Consum. Psychol. 26, 40–52. https:// ease-of-use. Inf. Manag. 34, 237–250. https://doi.org/10.1016/S0378-7206(98)
doi.org/10.1016/j.jcps.2015.03.001. 00096-2.
Dodds, W.B., Monroe, K.B., Grewal, D., 1991. Effects of price, brand, and store Kees, J., Berry, C., Burton, S., Sheehan, K., 2017. An analysis of data quality: professional
information on buyers’ product evaluations. J. Market. Res. 28, 307–319. https:// panels, student subject pools, and amazon’s mechanical Turk. J. Advert. 46,
doi.org/10.1177/002224379102800305. 141–155. https://doi.org/10.1080/00913367.2016.1269304.
eMarketer, 2020. Distribution of Digital Buyers in the United States by Age Group, 2020. Kefi, H., Maar, D., 2018. The power of lurking: assessing the online experience of luxury
Eroglu, S.A., Machleit, K.A., Davis, L.M., 2003. Empirical testing of a model of online brand fan page followers. J. Bus. Res. 117, 579–586. https://doi.org/10.1016/j.
store atmospherics and shopper responses. Psychol. Market. 20 https://doi.org/ jbusres.2018.08.012.
10.1002/mar.10064. Kim, J.B., 2015. The mediating role of presence on consumer intention to participate in a
Eroglu, S.A., Machleit, K.A., Davis, L.M., 2001. Atmospheric qualities of online retailing: social commerce site. J. Internet Commer. 4, 425–454. https://doi.org/10.1080/
a conceptual model and implications. J. Bus. Res. 54, 177–184. https://doi.org/ 15332861.2015.1092067.
10.1016/S0148-2963(99)00087-9. Kim, J.H., Kim, M., Yoo, J., Park, M., 2021. Consumer decision-making in a retail store:
Farivar, S., Turel, O., Yuan, Y., 2017. A trust-risk perspective on social commerce use: an the role of mental imagery and gender difference. Int. J. Retail Distrib. Manag. 49
examination of the biasing role of habit. Internet Res. 27 https://doi.org/10.1108/ https://doi.org/10.1108/IJRDM-10-2019-0353.
IntR-06-2016-0175. Kline, R.B., 2015. Principles and practices of structural equation modelling. In:
Fornell, C., Larcker, D.F., 1981. Structural equation models with unobservable variables Methodology in the Social Sciences, fourth ed. The Guilford Press, New York, NY,
and measurement error: algebra and statistics. J. Market. Res. 18, 382–388. https:// USA.
doi.org/10.2307/3150980. Kumar, N., Benbasat, I., 2002. Para-social presence and communication capabilities of a
Fortin, D.R., Dholakia, R.R., 2005. Interactivity and vividness effects on social presence web site: a theoretical perspective. e Serv. J. 1, 5–24. https://doi.org/10.2979/
and involvement with a web-based advertisement. J. Bus. Res. 58, 387–396. https:// esj.2002.1.3.5.
doi.org/10.1016/S0148-2963(03)00106-1. Labrecque, L.I., 2020. Color research in marketing: theoretical and technical
Friedrich, T., Schlauderer, S., Overhage, S., 2019. The impact of social commerce feature considerations for conducting rigorous and impactful color research. Psychol.
richness on website stickiness through cognitive and affective factors: an Market. https://doi.org/10.1002/mar.21359.
experimental study. Electron. Commer. Res. Appl. 36 https://doi.org/10.1016/j. Lee, E.J., Shin, S.Y., 2014. When do consumers buy online product reviews? Effects of
elerap.2019.100861. review quality, product type, and reviewer’s photo. Comput. Hum. Behav. https://
Gan, C., Wang, W., 2017. The influence of perceived value on purchase intention in doi.org/10.1016/j.chb.2013.10.050.
social commerce context. Internet Res. 27 https://doi.org/10.1108/IntR-06-2016- Lee, J.E., Watkins, B., 2016. YouTube vloggers’ influence on consumer luxury brand
0164. perceptions and intentions. J. Bus. Res. 69, 5753–5760. https://doi.org/10.1016/J.
Gavilan, D., Avello, M., Abril, C., 2014. The mediating role of mental imagery in mobile JBUSRES.2016.04.171.
advertising. Int. J. Inf. Manag. 34, 457–464. https://doi.org/10.1016/j. Lee, K.Y., 2012. Consumer processing of virtual experience in e-commerce: a test of an
ijinfomgt.2014.04.004. integrated framework. Comput. Hum. Behav. 28, 2134–2142. https://doi.org/
Gefen, D., Straub, D.W., 2004. Consumer trust in B2C e-Commerce and the importance of 10.1016/j.chb.2012.06.018.
social presence: experiments in e-Products and e-Services. Omega 32, 407–424. Lee, W., Xiong, L., Hu, C., 2012. The effect of Facebook users’ arousal and valence on
https://doi.org/10.1016/j.omega.2004.01.006. intention to go to the festival: applying an extension of the technology acceptance
Guo, Q., Zeng, Q., Zhang, L., 2022. What social factors influence learners’ continuous model. Int. J. Hospit. Manag. 31 https://doi.org/10.1016/j.ijhm.2011.09.018.
intention in online learning? A social presence perspective. Inform. Technol. People Leung, W.K.S., Shi, S., Chow, W.S., 2020. Impacts of user interactions on trust
ahead-of-print. https://doi.org/10.1108/ITP-02-2021-0151/FULL/XML. development in C2C social commerce: the central role of reciprocity. Internet Res. 30
Han, H., Xu, H., Chen, H., 2018. Social commerce: a systematic review and data https://doi.org/10.1108/INTR-09-2018-0413.
synthesis. Electron. Commer. Res. Appl. 30, 38–50. https://doi.org/10.1016/j. Li, C.Y., 2019. How social commerce constructs influence customers’ social shopping
elerap.2018.05.005. intention? An empirical study of a social commerce website. Technol. Forecast. Soc.
Hassanein, K., Head, M., 2007. Manipulating perceived social presence through the web Change 144. https://doi.org/10.1016/j.techfore.2017.11.026.
interface and its impact on attitude towards online shopping. Int. J. Hum. Comput. Li, X., Wang, M., Chen, Y., 2014. The impact of product photo on online consumer
Stud. 65, 689–708. https://doi.org/10.1016/j.ijhcs.2006.11.018. purchase intention: an image-processing enabled empirical study. In: Proceedings -
Hauser, D.J., Ellsworth, P.C., Gonzalez, R., 2018. Are manipulation checks necessary? Pacific Asia Conference on Information Systems. PACIS, 2014.
Front. Psychol. 9, 998. https://doi.org/10.3389/fpsyg.2018.00998. Li, Y., Xie, Y., 2020. Is a picture worth a thousand words? An empirical study of image
Hayes, A.F., 2018. Introduction to Mediation, Moderation, and Conditional Process content and social media engagement. J. Market. Res. 57, 1–19. https://doi.org/
Analysis, Second Edition: A Regression-Based Approach. the Guilford Press. 10.1177/0022243719881113.
Herzallah, D., Leiva, F.M., Liébana-Cabanillas, F., 2022. To buy or not to buy, that is the Liang, T.-P., Turban, E., 2011. Introduction to the special issue social commerce: a
question: understanding the determinants of the urge to buy impulsively on research framework for social commerce. Int. J. Electron. Commer. 16, 5–14.
Instagram Commerce. J. Res. Indian Med. 16 https://doi.org/10.1108/JRIM-05- https://doi.org/10.2753/JEC1086-4415160201.
2021-0145. Liang, T.-P.T.-P. a, Ho, Y.-T.Y.-T. b, Li, Y.-W. c Y.-W., Turban, E.E. d, 2011. What drives
Hess, T., Fuller, M., Campbell, D., 2009. Designing interfaces with social presence : using social commerce: the role of social support and relationship quality. Int. J. Electron.
vividness and extraversion to create social recommendation agents. J. Assoc. Inf. Commer. 16, 69–90. https://doi.org/10.2753/JEC1086-4415160204.
Syst. Online 10, 889–919. https://doi.org/10.1126/science.1063522. Limayem, M., Khalifa, M., Frini, A., 2000. What makes consumers buy from Internet? A
Hollebeek, L.D., Clark, M.K., Macky, K., 2020. Demystifying consumer digital cocreated longitudinal study of online shopping. IEEE Trans. Syst. Man Cybern. Syst. Hum. 30
value: social presence theory-informed framework and propositions. Rech. Appl. https://doi.org/10.1109/3468.852436.
Market. https://doi.org/10.1177/2051570720961986. Liu, P., Li, M., Dai, D., Guo, L., 2021. The effects of social commerce environmental
Hou, L., Pan, X., 2023. Aesthetics of hotel photos and its impact on consumer characteristics on customers’ purchase intentions: the chain mediating effect of
engagement: a computer vision approach. Tourism Manag. 94, 104653 https://doi. customer-to-customer interaction and customer-perceived value. Electron. Commer.
org/10.1016/J.TOURMAN.2022.104653. Res. Appl. 48 https://doi.org/10.1016/j.elerap.2021.101073.
Hu, N., Pavlou, P.A., Zhang, J., 2017. On self-selection biases in online product reviews. Lu, B., Fan, W., Zhou, M., 2016. Social presence, trust, and social commerce purchase
MIS Q. 41, 449–471. https://doi.org/10.25300/MISQ/2017/41.2.06. intention: an empirical research. Comput. Hum. Behav. 56, 225–237. https://doi.
Huang, Z., Benyoucef, M., 2017. The effects of social commerce design on consumer org/10.1016/j.chb.2015.11.057.
purchase decision-making: an empirical study. Electron. Commer. Res. Appl. 25, Ma, X., Meng, X., 2021. Image position and layout effects on user engagement of multi-
40–58. https://doi.org/10.1016/j.elerap.2017.08.003. image tweets. Proc. Assoc. Inform. Sci. Technol. 58 https://doi.org/10.1002/
Hulland, J., Miller, J., 2018. Keep on turkin. J. Acad. Market. Sci. 46, 789–794. https:// pra2.484.
doi.org/10.1007/s11747-018-0587-4. Mai, R., Niemand, T., Kraus, S., 2021. A tailored-fit model evaluation strategy for better
Itzchakov, G., Amar, M., Van Harreveld, F., 2020. Don’t let the facts ruin a good story: decisions about structural equation models. Technol. Forecast. Soc. Change 173,
the effect of vivid reviews on attitude ambivalence and its coping mechanisms. 121142. https://doi.org/10.1016/J.TECHFORE.2021.121142.
J. Exp. Soc. Psychol. https://doi.org/10.1016/j.jesp.2019.103938. Marks, D.F., 1973. Visual mental imagery in the recall of pictures. Br. J. Psychol. 64,
Jiang, L., Alice), Yang, Z., Jun, M., 2013. Measuring consumer perceptions of online 17–24.
shopping convenience. J. Serv. Manag. 24, 191–214. https://doi.org/10.1108/ Meeker, M., 2016. Internet Trends 2016—Code Conference. Kleiner Perkins.
09564231311323962/FULL/XML. Mehrabian, A., Russell, J.A., 1974. An Approach to Environmental psychology., an
Jin, S.V., Ryu, E., 2020. The roles of envy toward and parasocial interaction with Approach to Environmental Psychology. The MIT Press, Cambridge, MA, USA.
influencers in Instagram celebrity-based brand endorsement and social commerce. Miller, D.W., Hadjimarcou, J., Miciak, A., 2000. A scale for measuring advertisement-
J. Retailing Consum. Serv. 55, 102121 https://doi.org/10.1016/J. evoked mental imagery. J. Market. Commun. 6 https://doi.org/10.1080/
JRETCONSER.2020.102121. I’ll buy what she’s #wearing. 135272600345525.
Kanani, R., Glavee-Geo, R., 2021. Breaking the uncertainty barrier in social commerce: Mollen, A., Wilson, H., 2010. Engagement, telepresence and interactivity in online
the relevance of seller and customer-based signals. Electron. Commer. Res. Appl. 48 consumer experience: reconciling scholastic and managerial perspectives. J. Bus.
https://doi.org/10.1016/j.elerap.2021.101059. Res. 63, 919–925. https://doi.org/10.1016/j.jbusres.2009.05.014.
Kang, J.Y.M., Johnson, K.K.P., Wu, J., 2014. Consumer style inventory and intent to Mou, J., Benyoucef, M., 2021. Consumer behavior in social commerce: results from a
social shop online for apparel using social networking sites. J. Fash. Mark. Manag. 18 meta-analysis. Technol. Forecast. Soc. Change 167. https://doi.org/10.1016/j.
https://doi.org/10.1108/JFMM-09-2012-0057. techfore.2021.120734.
Nadeem, W., Khani, A.H., Schultz, C.D., Adam, N.A., Attar, R.W., Hajli, N., 2020. How
social presence drives commitment and loyalty with online brand communities? the

14
E.E. Vazquez et al. Journal of Retailing and Consumer Services 74 (2023) 103415

role of social commerce trust. J. Retailing Consum. Serv. 55, 102136 https://doi. Taylor, D.G., Lewin, J.E., Strutton, D., 2011. Friends, fans, and followers: do ads work on
org/10.1016/J.JRETCONSER.2020.102136. social networks? How gender and age shape receptivity. J. Advert. Res. 51, 258.
Niemand, T., Mai, R., 2018. Flexible cutoff values for fit indices in the evaluation of https://doi.org/10.2501/JAR-51-1-258-275.
structural equation models. J. Acad. Market. Sci. 46 https://doi.org/10.1007/ Toteva, I.T., Lutz, R.J., Shaw, E.H., 2021. The curious case of productivity orientation:
s11747-018-0602-9. the influence of advertising stimuli on affect and preference for subscription boxes.
Nowak, K.L., Biocca, F., 2003. The effect of the agency and anthropomorphism on users’ J. Retailing Consum. Serv. 63, 102677 https://doi.org/10.1016/J.
sense of telepresence, copresence, and social presence in virtual environments. JRETCONSER.2021.102677.
Presence Teleoperators Virtual Environ. 12, 481–494. https://doi.org/10.1162/ Turley, L.W., Milliman, R.E., 2000. Atmospheric effects on shopping behavior: a review
105474603322761289. of the experimental evidence. J. Bus. Res. 49, 193–211. https://doi.org/10.1016/
Olbrich, R., Holsing, C., 2011. Modeling consumer purchasing behavior in social S0148-2963(99)00010-7.
shopping communities with clickstream data. Int. J. Electron. Commer. 16 https:// Vazquez, E.E., 2021. Effect of an e-retailer’s product category and social media platform
doi.org/10.2753/JEC1086-4415160202. selection on perceived quality of e-retail products. Electron. Mark. 31, 139–157.
Paivio, A., 2013. Imagery and verbal processes. In: Imagery and Verbal Processes, first https://doi.org/10.1007/s12525-020-00394-8.
ed. Psychology Press, New York, USA. https://doi.org/10.4324/9781315798868. Vazquez, E.E., 2020. Effects of enduring involvement and perceived content vividness on
Paivio, A., 2008. Mental representations: a dual coding approach, mental digital engagement. J. Res. Interact. Mark. 14, 1–16. https://doi.org/10.1108/JRIM-
representations: a dual coding approach. https://doi.org/10.1093/acprof:oso/97801 05-2018-0071.
95066661.001.0001. Venturebeat.com, 2016. Facebook: 60 million businesses have Pages, 4 million actively
Paivio, A., 1991. Dual coding theory: retrospect and current status. Can. J. Psychol./Rev. advertise [WWW Document], 4.3.23. https://venturebeat.com/2016/09/27/face
Canad. Psychol. 45 https://doi.org/10.1037/h0084295. book-60-million-businesses-have-pages-4-million-actively-advertise/.
Park, M., Im, H., Kim, H.Y., 2020. “You are too friendly!” The negative effects of social Voorhees, C.M., Brady, M.K., Calantone, R., Ramirez, E., 2016. Discriminant validity
media marketing on value perceptions of luxury fashion brands. J. Bus. Res. 117, testing in marketing: an analysis, causes for concern, and proposed remedies.
529–542. https://doi.org/10.1016/j.jbusres.2018.07.026. J. Acad. Market. Sci. 44, 119–134. https://doi.org/10.1007/s11747-015-0455-4.
Pechmann, C.(Connie), Connie), 2015. Social marketing research trends in consumer Walther, J.B., 1992. Interpersonal effects in computer-mediated interaction: a relational
psychology. Rech. Appl. Market. 30, 124–127. https://doi.org/10.1177/ perspective. Commun. Res. 19 https://doi.org/10.1177/009365092019001003.
2051570715598732. Wang, F., Xu, H., Hou, R., Zhu, Z., 2023. Designing marketing content for social
Peng, C., Kim, Y.G., 2014. Application of the stimuli-organism-response (S-O-R) commerce to drive consumer purchase behaviors: a perspective from speech act
framework to online shopping behavior. J. Internet Commer. 13, 159–176. https:// theory. J. Retailing Consum. Serv. 70, 103156 https://doi.org/10.1016/J.
doi.org/10.1080/15332861.2014.944437. JRETCONSER.2022.103156.
Richard, M.-O., 2005. Modeling the impact of internet atmospherics on surfer behavior. Wang, Y.J., Hernandez, M.D., Minor, M.S., 2010. Web aesthetics effects on perceived
J. Bus. Res. 58, 1632–1642. https://doi.org/10.1016/j.jbusres.2004.07.009. online service quality and satisfaction in an e-tail environment: the moderating role
Rönkkö, M., Cho, E., 2022. An updated guideline for assessing discriminant validity. of purchase task. J. Bus. Res. https://doi.org/10.1016/j.jbusres.2009.01.016.
Organ. Res. Methods. https://doi.org/10.1177/1094428120968614. Weisberg, J., Te’eni, D., Arman, L., 2011. Past purchase and intention to purchase in e-
Rundle-Thiele, S., 2015. Looking back and moving forwards: an agenda for social commerce: the mediation of social presence and trust. Internet Res. 21, 82–96.
marketing research. Rech. Appl. Market. https://doi.org/10.1177/ https://doi.org/10.1108/10662241111104893.
2051570715599338. Wham, E., 2018. What is social commerce: details and data [WWW document], 4.3.23.
Sabate, F., Berbegal-Mirabent, J., Cañabate, A., Lebherz, P.R., 2014. Factors influencing http://www.project-disco.org/innovation/091718-social-commerce-details-data.
popularity of branded content in Facebook fan pages. Eur. Manag. J. 32 https://doi. Wu, W., Wang, S., Ding, G., Mo, J., 2023. Elucidating trust-building sources in social
org/10.1016/j.emj.2014.05.001. shopping: a consumer cognitive and emotional trust perspective. J. Retailing
Saprikis, V., Avlogiaris, G., 2021. Modeling users’ acceptance of mobile social commerce: Consum. Serv. 71, 103217 https://doi.org/10.1016/J.JRETCONSER.2022.103217.
the case of ‘Instagram checkout.’ Electronic Commerce Research. https://doi. Wu, Y.L., Li, E.Y., 2018. Marketing mix, customer value, and customer loyalty in social
org/10.1007/s10660-021-09499-4. commerce: a stimulus-organism-response perspective. Internet Res. https://doi.org/
Sautter, P., Hyman, M.R., Lukošius, V., 2004. E-Tail atmospherics: a critique of the 10.1108/IntR-08-2016-0250.
literature and model extension. J. Electron. Commer. Res. 5, 14–24. Xu, F., Han, Z., Piao, J., Li, Y., 2019. Modelling the online purchase behavior in social E-
Senthil Kumar, S., Ramachandran, T., Panboli, S., 2015. Product recommendations over commerce. Proc ACM Hum. Comput. Interact. https://doi.org/10.1145/3359167. I
Facebook: the roles of influencing factors to induce online shopping. Asian Soc. Sci. think you’ll like it.
11 https://doi.org/10.5539/ass.v11n2p202. Yahia, I. Ben, Al-Neama, N., Kerbache, L., 2018. Investigating the drivers for social
Shen, K.N., Khalifa, M., 2008. Exploring multidimensional conceptualization of social commerce in social media platforms: importance of trust, social support and the
presence in the context of online communities. Int. J. Hum. Comput. Interact. 24, platform perceived usage. J. Retailing Consum. Serv. 41, 11–19. https://doi.org/
722–748. https://doi.org/10.1080/10447310802335789. 10.1016/J.JRETCONSER.2017.10.021.
Short, J., Williams, E., Christie, B., 1976. The Social Psychology of Telecommunications. Zhang, H., Lu, Y., Gupta, S., Zhao, L., 2014. What motivates customers to participate in
John Wiley & Sons, Ltd., Hoboken, NJ. social commerce? The impact of technological environments and virtual customer
Silverstein, M.J., Sayre, K., 2009. The female economy. Harv. Bus. Rev. 87. experiences. Inf. Manag. 51, 1017–1030. https://doi.org/10.1016/J.
Skalski, P., Tamborini, R., 2004. Social presence and information processing: an IM.2014.07.005.
application of the heuristic-systematic model of persuasion. In: Annual Conference of Yuan, Y., Xu, F., Cao, H., Zhang, G., Hui, P., Li, Y., Jin, D., 2021. Persuade to click:
the National Communication Association (Chicago, IL, USA). context-aware persuasion model for online textual advertisement. IEEE Trans.
Slama, M.E., Tashchian, A., 1987. Validating the S-O-R paradigm for consumer Knowl. Data Eng. 35 https://doi.org/10.1109/TKDE.2021.3110724.
involvement with a convenience good. Market. Sci. 15, 36–45. https://doi.org/ Zhang, K.Z.K., Benyoucef, M., 2016. Consumer behavior in social commerce: a literature
10.1007/BF02721952. review. Decis. Support Syst. 86 https://doi.org/10.1016/j.dss.2016.04.001.
Song, S., Park, S. “Brian”, Park, K., 2021. Thematic analysis of destination images for Zhang, M., Luo, N., 2016. Understanding relationship benefits from harmon ious brand
social media engagement marketing. Ind. Manag. Data Syst. 121, 1375–1397. community on social media. Internet Res. 26 https://doi.org/10.1108/IntR-05-
https://doi.org/10.1108/IMDS-12-2019-0667/FULL/XML. 2015-0149.
Song, S.S., Kim, M., 2012. Does more mean better? An examination of visual product Zhao, W., Hu, F., Wang, J., Shu, T., Xu, Y., 2023. A systematic literature review on social
presentation in e-retailing. J. Electron. Commer. Res. 13. commerce: assessing the past and guiding the future. Electron. Commer. Res. Appl.
Spangenberg, E.R., Crowley, A.E., Henderson, P.W., 1996. Improving the store 57, 101219 https://doi.org/10.1016/j.elerap.2022.101219.
environment: do olfactory cues affect evaluations and behaviors? J. Market. https://
doi.org/10.1177/002224299606000205.
Starke, A.D., Willemsen, M.C., Trattner, C., 2021. Nudging healthy choices in food search
Further reading
through visual attractiveness. Front Artif Intell. 4 https://doi.org/10.3389/
frai.2021.621743. Callejo, P., Gramaglia, M., Cuevas, R., Cuevas, Á., 2021. A deep dive into the accuracy of
Statista, 2019. Women’s & Girls’ Apparel Report 2019 (New York, United States). IP Geolocation Databases and its impact on online advertising. IEEE Trans. Mobile
Straub, D.W., 1994. The effect of culture on IT diffusion: E-mail and FAX in Japan and Comput. https://doi.org/10.1109/TMC.2022.3166785.
the U.S. Inf. Syst. Res. 5 https://doi.org/10.1287/isre.5.1.23.

15
Received: 6 September 2022 Revised: 30 August 2023 Accepted: 9 October 2023
DOI: 10.1002/cb.2274

RESEARCH ARTICLE

Female advertising models and their body mass index (BMI):

How does BMI similarity affect perceived product quality?

Erik Ernesto Vazquez 1 | Reto Felix 2 | Lorena Siliceo 3

1
Léonard De Vinci Pôle Universitaire, Research
Center, Paris La Défense, France Abstract
2
Robert C. Vackar College of Business and Recent research has pointed out the importance of appearance similarity as a special
Entrepreneurship, University of Texas Rio
Grande Valley, Edinburg, Texas, USA instance of homophily. The current paper introduces body mass index (BMI) similarity
3
Universidad Anáhuac México, Facultad de as a convenient and powerful proxy of perceived appearance homophily in the con-
Economía y Negocios, Huixquilucan, Estado de
text of female advertising models by showing process evidence and boundary condi-
México, Mexico
tions for the effect of BMI similarity on marketing-relevant downstream variables.
Correspondence
Study 1 tests how BMI similarity relates to a traditional measure of appearance
Erik Ernesto Vazquez, Léonard De Vinci Pôle
Universitaire, Research Center, Paris La homophily and perceived reliability of a female advertising model. Study 2 shows that
Défense 92916, France. BMI similarity influences perceived product quality and purchase intention. Website
Email: [email protected]
vividness negatively moderates the relationship between BMI similarity and product
quality. Study 3 tests for alternative explanations and provides support for the medi-
ating effect of appearance homophily for the relationship between BMI similarity and
perceived product quality. The findings provide marketing managers with important
insights on how to increase their marketing effectiveness by integrating BMI similar-
ity into their marketing communications. Additionally, using BMI similarity serves as
an alternative way to promote diversity and inclusion of models with plus-size body
type often sought by societal brands.

1 | I N T RO DU CT I O N which refers to a general assessment of appearance and which may

include different criteria such as facial features and properties of the
Similarity is a powerful construct for human relationships. People tend body (Bekk et al., 2017; Ladhari et al., 2020). It has been suggested
to be homophilous—that is, they usually develop stronger ties with that appearance similarity tends to be powerful in human relationships
other people who are similar to them (Borgatti & Halgin, 2011). Previ- because it allows people to infer socially relevant information about
ous research suggests that similarity between two persons increases others (Bjornsdottir et al., 2021). For example, Ladhari et al. (2020)
attraction (Rodrigues et al., 2017), communication effectiveness find that appearance homophily (defined as the extent to which a per-
(Thakor et al., 2008), perceived trust (Leonhardt et al., 2020), per- son perceives that another person looks similar to them) between a
ceived credibility (Sokolova & Kefi, 2020), and influence (Balabanis & video blogger (vlogger) and a viewer increases the vlogger's perceived
Chatzopoulou, 2019). In a consumption context, messages from con- popularity, which in turn increases purchase intention. In a similar
sumers who are perceived to be similar increase message persuasive- vein, Bekk et al. (2017) show that similarity in perceived attractiveness
ness and purchase intention (Prendergast et al., 2010). In online moderates the influence of endorser attractiveness on consumer
support groups, dissimilarity among group members impedes self- responses to advertising.
disclosure, which in turn weakens the social ties of such groups Despite the importance that appearance homophily has in a mar-
(Pechman et al., 2021; Vazquez, 2021). To summarize, the extant liter- keting context, a major drawback for providing managerially relevant
ature finds important effects of similarity in demographics, values, and opportunities for intervention is that appearance homophily implies a
beliefs for marketing-relevant downstream variables. However, recent subjective assessment of another person and hence depends on mea-
research has pointed out the importance of appearance similarity, surement scales to capture consumers' perceptions. For example,

J Consumer Behav. 2023;1–19. wileyonlinelibrary.com/journal/cb © 2023 John Wiley & Sons Ltd. 1
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 VAZQUEZ ET AL.

based on McCroskey et al.'s (2006) measure of homophily, Ladhari Furthermore, the present research introduces perceived website viv-
et al. (2020) use a 16-item scale to measure consumers' perceived idness as a moderator for the relationship between BMI similarity on
similarity with a vlogger in terms of attitudes, general background, perceived product quality in a retailing context for fashion apparel.
values, and appearance. Applying such scales is indeed feasible for Study 3 explores alternative explanations for our findings by testing
market research purposes, but marketers are usually not able to the mediating effects of appearance homophily, trustworthiness,
receive appearance-related homophily information in a comprehen- expertise, familiarity, and likability on the relationship between BMI
sive way from all customers as long as this information is based on similarity and perceived product quality. A supplementary study
subjective measurement scales. (reported in the Web Appendix) on group similarity provides addi-
The current research addresses this issue in a novel way by dem- tional evidence for the importance of similarity measures for per-
onstrating that body mass index (BMI) functions satisfactorily as a ceived product quality in the context of an existing fashion clothing
proxy for perceived appearance homophily in a fashion retail setting. company (boohoo.com). Finally, we provide a general discussion of
Situating our research in the context of comparably thin female adver- our findings, including implications for academia and marketing man-
tising models allows us to avoid confounds related to the body size agers, and suggest avenues for future research.
(e.g., thin, average, plus-size) of the model and to explore the phenom-
enon under a theoretical perspective that builds on the prevalence of
thin models in contemporary advertising (Cohen et al., 2017; 2 | THEORETICAL FRAMEWORK
Volonté, 2019). Hence, our work complements research about female
portrayals of diverse body sizes and the positive effects of inclusive 2.1 | BMI similarity, appearance homophily,
advertising of plus-size models (e.g., Åkestam et al., 2017; Cinelli & and perceived reliability of the advertising model
Yang, 2016; Janssen & Paas, 2014; Joo & Wu, 2021; Pounders, 2018).
Furthermore, focusing on advertising models in combination with Homophily is a specific type of social contagion, that is, people's reli-
same-gender (i.e., women) participants in the present studies ance on information from others (Bilgicer et al., 2015). Early conceptu-
averts gender confounds (compare previous studies, e.g., Felix & alizations of homophily distinguished between status homophily and
Garza, 2012; Lv et al., 2022, who use female samples in their research value homophily (Lazarsfeld & Merton, 1954). Whereas status homo-
on body appearance). phily refers to formal, informal, or ascribed status (which includes
Using BMI rather than subjective measurement scales as a proxy sociodemographic characteristics such as ethnicity, gender, and age),
for appearance homophily has several advantages. First, it is relatively value homophily alludes to similarity based on values, beliefs, and atti-
easy to obtain BMI data because they are based on consumers' height tudes (McPherson et al., 2001). More recent research suggests that
and weight—two metrics that marketers may already have in their apart from status and values, similarity of appearance is an important
customer databases. Further, advances in face recognition software factor when predicting attitudes toward influencers and salespersons
and artificial intelligence now allow for a relatively reliable calculation (Bjornsdottir et al., 2021; Ladhari et al., 2020). However, the disadvan-
of BMI based on just a facial image (Dantcheva et al., 2018; Wen & tage of appearance similarity is that it is a subjective rather than
Guo, 2013). Hence, we argue that using BMI rather than traditional objective measure of similarity, and hence it depends on psycho-
measurement scales provides marketers with important opportunities graphic measurement scales to capture how similar a customer feels
to predict managerially relevant outcome variables, such as perceived in comparison to a salesperson or brand endorser. Table 1 provides an
product quality and purchase intention. overview of selected studies on homophily and source similarity in
We also include an important boundary condition, website vivid- marketing and consumer research. As can be seen from the table, pre-
ness, for the effect of BMI on quality perception and purchase inten- vious studies on homophily relating to personality traits, values, or
tion in our research. Specifically, we observe that many marketers use appearance depend on subjective measures of homophily, except for
elements such as animations, dynamic website objects, sound, and demographic similarity based on variables such as age or gender.
action buttons to enhance the vividness of their websites and thereby Importantly, similarity in appearance is commonly measured through
capture consumers' attention. Drawing on cognitive depletion theory self-reported scales (Ladhari et al., 2020; Torres, 2007). Addressing
(e.g., Hildebrand et al., 2021; Sokolova & Krishna, 2016), we expect this issue, the current research suggests that similarity based on BMI
that the use of such elements may decrease the salience of may act as a parsimonious alternative for appearance homophily.
appearance-related information and attenuate the relationship Traditionally, marketers have postulated that extremely thin
between BMI similarity and perceived product quality. models with a very low BMI convey ideals of beauty and aesthetics
Our paper is structured as follows: We first provide the theoreti- (Venkatesh et al., 2010) and draw consumer purchases through aspira-
cal background for our studies and comment on our hypotheses. Next, tional values (Barry, 2014). The rationale for promoting an extremely
Study 1 tests our assumption that BMI similarity is positively related thin body image that is unattainable for most women is grounded in
with appearance homophily, which in turn is positively associated with the expectation that unfulfilled aspirations secure continued pur-
the perceived reliability toward a fashion model. In addition, Study chases from customers hunting their appearance-based ideals
1 excludes perceived originality as an alternative explanation for this (Borland & Akram, 2007). Thus, it is not surprising that women with
effect. Study 2 demonstrates the predicted association between BMI high BMI are typically less satisfied with their body image and desire
similarity and perceived product quality/purchase intention. to be thinner (D'Alessandro & Bill Chitty, 2011). The complexity of
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 3

TABLE 1 Selective literature overview of homophily measures in marketing and consumer research.

Measurement of
homophily
(subjective vs.
Authors Journal Homophily definition Homophily domain objective) Operationalization
Balabanis and Psychology & Communication Similarity in social Subjective Four items of perceived subjective
Chatzopoulou Marketing volume is higher class, background, similarity to a blogger.
(2019) when the source and status
and the recipient
are similar.
Bekk et al. (2017) Marketing The notion that Similarity in physical Subjective Comparison of subjectively
Letters similarity leads to attractiveness measured physical
liking and attractiveness for an advertising
attraction, whereas endorser and one's own
dissimilarity leads attractiveness.
to repulsion.
Bilgicer et al. Journal of The principle that a Similarity based on Objective Clustering approach based on
(2015) Retailing contact between demographic demographic variables.
similar people variables, such as
occurs at a higher income and
rate than among education
dissimilar people.
Campbell et al. Journal of The idea that the Similarity based on Subjective Comparison of subjectively
(1988) Marketing more similar the self-esteem, values, measured self-esteem, values,
parties in a dyad and and interorganizational contact
are, the more likely interorganizational between buyers and sellers.
is a favorable contact
outcome.
De Bruyn and International Perceptual homophily Perceptual homophily Subjective Respondents' perceptions of
Lilien (2008) Journal of is defined as values, likes and dislikes and
Research in similarities experiences regarding an
Marketing between two acquaintance.
people's values,
likes, dislikes, and
experience.
Fu et al. (2019) Journal of External similarity External and internal Objective and External similarity: Respondents'
Strategic refers to similarity subjective information on name, age,
Marketing characteristics that birthday, and place of residence
can be obtained of other people in an online
without deep community. Internal similarity:
interaction. Internal Respondents' perception
similarity refers to regarding values, interests,
characteristics that attitudes, and preferences of
can only be other people in an online
obtained through community
deep interaction.
Ladhari et al. Journal of The degree to which Similarity in attitudes, Subjective Sixteen-item scale measuring
(2020) Retailing and people who background, value, perceived similarity to a vlogger
Consumer interact are similar and appearance based on attitude, background,
Services in beliefs, value, and appearance.
education, social
status and the like.
Leonhardt et al. Journal of Extent to which Similarity in attitudes Subjective Four items referring to similar
(2020) Business individuals are and behaviors attitudes and behaviors as
Research similar to others. compared to a member of a
Facebook group.
Pechman et al. Journal of The tendency of Demographic Objective Dyadic dissimilarity index based
(2021) Consumer interpersonal similarity on observed demographics
Psychology similarities to breed (gender, age, marital status,
social connection. employment status) between
two people in a support group.

(Continues)
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 VAZQUEZ ET AL.

TABLE 1 (Continued)

Measurement of
homophily
(subjective vs.
Authors Journal Homophily definition Homophily domain objective) Operationalization
Prendergast et al. International The degree to which Similarity in interests Subjective Four items measuring similarity of
(2010) Journal of individuals are interests between respondent
Advertising similar in terms of and an online forum.
certain shared
social
characteristics
Risselada et al. Journal of The principle that a Demographic Objective Similarity score between
(2014) Marketing contact between similarity customers based on age,
similar people gender, education level, and
occurs at a higher income for the adoption of an
rate than among innovative product.
dissimilar people.
Rodrigues et al. Computers in The degree of Attitude homophily Subjective Three items referring to similar
(2017) Human perceived (rather ideas, attitudes, and beliefs as
Behavior than actual) compared to unfamiliar people.
similarity with
another person.
Sokolova and Kefi Journal of The degree to which Attitude homophily Subjective Two items referring to similar
(2020) Retailing and people tend to thinking and values as
Consumer bond with similar compared to an influencer.
Services others.
Streukens and Psychology & Preference of people Similarity in Subjective Comparison of subjectively
Andreassen Marketing who are similar to conscientiousness, measured personality traits for
(2013) oneself, leading to extraversion, customers and frontline
attraction. agreeableness, and employees.
stability
Thakor et al. Journal of Common meanings, Age similarity as a Subjective Estimation of reference
(2008) Retailing attitudes, or beliefs proxy for consumers' age in a service
lead to increased stereotyping encounter.
and more effective
interaction.
Torres (2007) Journal of The degree to which Multi-dimensional Subjective Twenty-seven items measuring
Business pairs of individuals perceived perceived similarity to
Research are similar with homophily advertising models based on
respect to attitudes, beliefs, personality,
attributes such as and appearance.
beliefs, values,
education, and
socials status.
van Esch et al. Marketing A tendency for Perceived general Subjective Three items measuring perceived
(2018) Intelligence & consumers to similarity general similarity to a blogger.
Planning gravitate toward
and form
relationships with
those who are
similar to them in
values, lifestyle,
gender, education
and other
attributes.
Von Wangenheim European The degree to which Similarity in Subjective Two items measuring

n
and Bayo Journal of individuals are preferences and respondents' perceived
(2004) Marketing similar in terms of values. similarity in terms of
certain attributes. preferences and values to a
consumer providing word-
of-mouth communication.
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 5

BMI effects on marketing-relevant downstream variables is further perceived reliability of a message source is an important component
demonstrated by findings which suggest that average advertising of trust (Soh et al., 2009), we hypothesize the following:
models, as compared to thin models, may attenuate consumers' body
anxiety, but at the same time may have negative effects on brand atti- Hypothesis 2. Perceived appearance homophily is posi-
tudes and purchase intention through a competing path that triggers tively associated with perceived reliability of the adver-
repulsion of natural models (Borau & Bonnefon, 2017). tising model.
The current research reconciles the inconclusive and in part con-
flicting results from previous studies on BMI by suggesting that BMI
difference between a customer and an endorser, rather than the abso- 2.2 | MI similarity, perceived product quality,
lute BMI of the endorser, provides meaningful insights for marketing and purchase intention
theory and practice. For example, rather than arguing that “moder-
ately thin advertising models are optimal, most of the time” H1 in Study 1 establishes our baseline assumption of a positive rela-
(Janssen & Paas, 2014, p. 167), we posit that smaller differences tionship between BMI similarity and appearance homophily. Further-
between the customer's and the advertising model's BMI lead to more more, H2 predicts a positive indirect relationship between BMI
positive marketing outcomes, such as attitudes toward the advertising similarity and perceived reliability of the advertising model. The out-
model and purchase intention. We test this proposition in the context come variable of perceived reliability of the advertising model is rele-
of thin female advertising models, which allows us to avoid confounds vant for marketers but does not necessarily translate into higher
based on the size (e.g., thin, average, or plus-size) and gender of the performance for the brand. Building on this limitation from H2,
advertising model. Furthermore, we intentionally focus on thin female hypotheses H3 and H4 in this section posit a more immediate associa-
advertising models because even though average and plus-size models tion between BMI similarity and two performance-based marketing
have become more popular in recent years (Cinelli & Yang, 2016; metrics: perceived product quality and purchase intention.
Pounders, 2018), the female thin-ideal still prevails in contemporary We focus on perceived product quality in Studies 2 and 3 since
advertising (Cohen et al., 2017; Volonté, 2019). Hence, we hypothe- the main purpose of employing an advertising model in the context of
size and empirically test in Study 1 that BMI similarity toward thin fashion apparel is indeed the promotion of the product the company
advertising models is positively associated with a conventional mea- is trying to sell, and perceived product quality is one of the main
sure of appearance-based homophily. drivers of purchase intention (Bagozzi & Khoshnevis, 2022). For con-
sumer behavior and decision making in marketing, understanding per-
Hypothesis 1. BMI similarity is positively associated ceptions customers have regarding the quality of products is usually
with perceived appearance homophily. more relevant than (a) manufacturers' claims about product quality or
(b) objective measures of product quality based on engineering stan-
We further posit that appearance homophily relates positively to dards (Stone-Romero et al., 1997). Following Golder et al. (2012), we
how reliable the advertising model is perceived.1 Similar to the reliabil- conceptualize perceived product quality as a conversion of perceived
ity of influencers (Filieri et al., 2023) and salespersons (Homburg & attributes into an aggregated evaluation of quality. Drawing on
Stock, 2005), we argue that the perceived reliability of advertising insights from source attractiveness model (McCracken, 1989;
models is crucial for marketers because it is an important component McGuire, 1985; Montoya & Horton, 2013), we predict that a high sim-
of source credibility and trust (Soh et al., 2009). Our hypothesis draws ilarity between the BMI of the customer and the advertising model
on the source attractiveness model (McGuire, 1985), which identifies will not only affect perceived traits of the endorser, but also extend to
familiarity, likability, and similarity to the source of a message as the the products that are promoted by the advertising model. This predic-
three main predictors of message effectiveness (McCracken, 1989). tion is based on the observation that consumers take peripheral cues
This positive effect of similarity between the source and the receiver (such as the attractiveness of the source triggered by the similarity to
of a message on attraction has also been emphasized by Byrne (1971) the advertising model) as indicators for product quality (Petty &
in his similarity-attraction theory. The underlying psychological mech- Cacioppo, 1986). Similar spillover effects have been shown in the con-
anism for the similarity-attraction effect contends that individuals text of traditional homophily measures for attitudes toward the prod-
favor stimuli that reinforce the logic and consistency of their world, uct (Bekk et al., 2017; Thakor et al., 2008), as well as for purchase
and that people who are similar to us generate such a reinforcement intention (Ismagilova et al., 2020; Lou et al., 2019).
(Montoya & Horton, 2013). Hence, similarity between a source and
receiver of a message results in higher levels of attractiveness, with Hypothesis 3. BMI similarity is positively associated
positive effects on trust (Kim & Kim, 2021; Racherla et al., 2012) and with perceived product quality of the apparel promoted
a variety of other outcome variables, such as message effectiveness, by the advertising model.
credibility, and product perceptions (Ertug et al., 2022). Given that
In addition, Study 2 also tests whether the association between
1
Congruent with prior literature (Bekk et al., 2017; Bjornsdottir et al., 2021; Ladhari
BMI similarity and perceived product quality extends to purchase
et al., 2020), we conceptualize appearance homophily as one form of similarity between a
reference person and a consumer. intention for the apparel product promoted by the advertising model.
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 VAZQUEZ ET AL.

Including purchase intention into our framework allows us to test the effect of a message because these elements distract the viewer. Thus,
association of BMI similarity with a variable that is of direct economic vividness in communication can backfire and distract from other ele-
relevance to managers. The positive relationship between perceived ments of a communication message (Guadagno et al., 2011). Hence, in
product quality and purchase intention has been demonstrated in the the context of the current research, we expect that vivid elements on
extant literature (Bagozzi & Khoshnevis, 2022; Das Guru & a website (e.g., animations and additional sounds) will distract the
Paulssen, 2020). Hence, viewer's attention and interfere with the processing of other informa-
tion, such as the comparison of the viewer's own with the advertising
Hypothesis 4. Perceived product quality of the apparel model's BMI. We posit that cognitive depletion theory (Sokolova &
promoted by the advertising model is positively associ- Krishna, 2016) can explain this effect. Specifically, cognitive depletion
ated with purchase intention. theory suggests that tasks that draw consumers' attention will require
them to assign cognitive resources to this task, which in turn
Based on the hypothesized associations between BMI similarity, decreases their propensity to assign cognitive resources to other tasks
perceived appearance homophily, and perceived product quality, we (Hildebrand et al., 2021). Consequently, when viewers' attention is
further postulate that perceived appearance homophily mediates the drawn toward vivid message elements, such as animations, sounds,
relationship between BMI similarity and perceived product quality. flashing fonts, etc., more cognitive capacity is assigned to the proces-
Specifically, we posit that BMI similarity has a positive relationship sing of the vivid message elements and less cognitive capacity is
with perceived appearance homophily (as stated in H1), and we assigned to other message elements. Hence,
expect that perceived appearance homophily is positively associated
with perceived product quality. We investigate this mediating effect Hypothesis 6. Website vividness negatively moderates
of perceived appearance homophily in Study 3, where we also test for the relationship between BMI similarity and perceived
alternative explanations. Hence, product quality.

Hypothesis 5. Perceived appearance homophily posi-

tively mediates the relationship between BMI similarity 3 | ST UDY 1
and perceived product quality.
Study 1 explores the effects of BMI similarity on consumers' percep-
tions of appearance homophily and their subsequent perceptions of reli-
2.3 | Website vividness ability toward a human advertising model (H1 and H2) in the context of
an online apparel retail store. To address a potential alternative explana-
Vividness has been defined as the “ability of a technology to produce tion for our hypothesized effect of BMI similarity, we include con-
a sensorially rich mediated environment” (Steuer, 1992, p. 80). The sumers' perceptions of originality as a second mediator in our model.
effect of a vivid (as opposed to a less vivid) communication message
on downstream variables such as persuasion and message liking typi-
cally depends on whether the vivid message is more (vs. less) resource 3.1 | Method
demanding than the less vivid message (Keller & Block, 1997;
Vazquez, 2020). For example, Coyle and Thorson (2001) manipulate 3.1.1 | Participants and procedure
vividness of a commercial website through the addition of audio
effects and animation, and they report higher levels of telepresence Two-hundred and three women (Mage = 38.6, SDage = 12.8) were
and more positive and enduring attitudes for consumers in the high recruited from Amazon Mechanical Turk (MTurk) to take part in an
vividness condition. In a similar vein, Fortin and Dholakia (2005) show online study in exchange for a small cash incentive following recent
that for a web-based advertisement, higher levels of vividness related studies (e.g., Vazquez, 2021). The majority of the participants
increase social presence and involvement, which in turn positively (59.4%) had an annual household income of $50,000 or above. Further,
influence attitudes toward the ad, attitudes toward the brand, and 66.7% of the sample had an associate degree or higher level of educa-
purchase intention. Yousaf et al. (2020) report that greater message tion. Participants first saw an image of a human advertising model
vividness (manipulated through interactive/audio-visual content) on (21 years of age, height 50 600 , 110.2 lbs, BMI of 17.8) (Appendix A,
social media platforms leads to higher customer engagement. In sum- Figure A1). Hence, our advertising model had a substantially lower BMI
mary, these studies find that vividness facilitates consumer informa- than the average US female population which is around 28.42 (Centers
tion processing. However, if a vivid communication message requires for Disease Control and Prevention, 2020). Our use of a comparably
additional cognitive resources and attention, the effect of the vivid skinny advertising model is consistent with the predominant tendency
elements in the message will most likely be distracting and have a in the industry to use skinny advertising models since they are fre-
negative effect on marketing-relevant downstream variables. For quently deemed to be culturally more attractive (Lin et al., 2018;
example, Frey and Eagly (1993) find that vivid elements, such as color- Semaan et al., 2018). We acknowledge that using only one (comparably
ful language or picturesque examples, can undermine the persuasive skinny) advertising model does not allow us to claim that the effects of
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 7

BMI similarity on downstream variables will also hold with a regular extant literature. For example, Li and Peng (2021) report a positive
or overweight advertising model. However, showing the same ad- relationship of r = .37 (p < .001) between homophily (i.e., the degree
vertising model to all participants allows us to exclude confounds of similarity between a consumer and a social media influencer) and
(e.g., differences in perceived attractiveness) that would occur when perceived originality of the influencer. Furthermore, prior research
using different advertising models. Next, participants were asked to suggests that perceived originality conveys an aspirational desire in
respond to items relating to their perceptions of the advertising model. apparel markets (Fratto et al., 2006; Zebal & Jackson, 2019) that may
These items measured the latent constructs of interest which were ran- translate into positive attitudes toward the advertising model. We
domly shown to participants to avoid priming effects. measured perceived originality with a five-item semantic differential
originally developed by O'Quin and Besemer (1989). This scale mea-
sures the degree to which an advertisement, sales presentation, sales
3.1.2 | Measures promotion, or event marketing communication is viewed as being
original (Heiser et al., 2008; White & Smith, 2001). Since the human
The dependent variable—perceived reliability of the advertising model— model used in our study also portrayed a specific age and ethnicity
was measured with an adapted version of the advertisement trust- (21 years, white), we included the similarity between participants' eth-
reliability scale from Soh et al. (2009). This nine-item Likert scale mea- nicity (Torres, 2007) and age (Dwyer et al., 1998) as control variables.
sures the degree to which a person believes that an advertisement is Specifically, we calculated the following two similarity indexes. For
truthful and accurate. The mediator variable—perceived appearance age, we subtracted the absolute age difference between the human
homophily—was measured with the scale originally developed by model and each participant's age from 100. Hence, smaller age differ-
McCroskey et al. (1975). This four-item bipolar scale measures the ences result in a higher number on the index, which represents higher
degree to which a person perceives appearance homophily and interper- levels of age similarity. For ethnicity, we generated a dummy variable
sonal similarity in human communication. It has been widely used in pre- that was coded 1 when the human model and participant had the
vious research in marketing and apparel contexts (Hudders & De same ethnicity or 0 otherwise. Appendix A, Table A1, shows the mea-
Jans, 2022; Ladhari et al., 2020) as well as health and human communica- surement items and reliability assessment for Study 1. The psycho-
tion research (Navarro et al., 2022; van der Land et al., 2015). The inde- metric properties of our measurement scales were satisfying.
pendent variable—BMI similarity—was calculated by comparing the BMI Specifically, the factor loadings exceeded the recommended threshold
of the advertising model with the BMI of each participant. Specifically, of 0.6 for each item and the composite reliability scores for each
participants indicated their height and weight, which in turn allowed us latent construct were beyond the recommended minimum value of
to determine each participants' BMI (Appendix A, Figure A2 shows a dis- 0.7, confirming the reliability of the measures (Bagozzi & Yi, 1988).
tribution of BMI of each participant). To ensure that higher values on our Each latent construct had an AVE above the minimum recommended
BMI-based similarity measure indicated higher (as opposed to smaller) value of 0.5 (MacKenzie et al., 2011) Furthermore, all AVE values
levels of similarity, we constructed an index of BMI similarity by subtract- were larger than the highest squared correlation between different
ing the absolute difference between the BMI of the advertising model latent constructs, thus confirming the discriminant validity of the mea-
and the BMI of the participant from 100. For example, for a participant sures (Fornell & Larcker, 1981; Voorhees et al., 2016). The descriptive
with a BMI of 24.5, the absolute BMI difference would be statistics and correlation matrix for Study 1 are shown in Table 2.
24.5  17.8 = 6.7, and the resulting BMI similarity index would be
100  6.7 = 93.3. This procedure resulted in a maximum value of
100 for the BMI similarity index when participants' BMI was identical to 3.2 | Results
the BMI of 17.8 for the advertising model appearing in the online store.
To test H1 and H2, we used a parallel mediation model which evalu-
ates the relationship between BMI similarity and perceived reliability
3.1.3 | Controlling for alternative explanations of the advertising model mediated by (1) perceived appearance
homophily and (2) perceived originality. Tests for significance and the
To address a potential alternative explanation for the relationship estimation of confidence intervals at 95% followed the 10,000 bias-
between BMI similarity and perceived reliability of the advertising corrected bootstrap samples procedure as recommended by Hayes
model, we included perceived originality of the advertising model as a (2022). All analyses were conducted using the processR package of
second mediator into our model. Originality has been defined as the the statistical software R (Moon & Hong, 2021). Figure 1 shows the
degree of newness, innovation, sophistication, and differentiation of a standardized path coefficients for this model.

et al., 2020). Because a source that
source, action, or product (Casalo Most importantly, BMI similarity was significantly and positively
is closer to oneself should be perceived as more elaborated, sophisti- related with perceived appearance homophily (0.373, p < .001). Further,
cated, and unique, we posit that consumers will feel that a source or perceived appearance homophily had a positive association with per-
stimulus is more original if the source is more similar to how they are ceived reliability of the advertising model (0.250, p < .001). These find-
or perceive themselves. Preliminary evidence for a positive relation- ings support H1 and H2. On the other hand, the relationship between
ship between similarity and perceived originality can be found in the BMI similarity and perceived originality was not statistically significant
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 VAZQUEZ ET AL.

TABLE 2 Descriptive statistics and correlations for Study 1.

Variable M SD 1. 2. 3. 4. 5.
1. Perceived reliability 4.31 1.24 -
2. Perceived appearance homophily 3.25 1.64 0.30*** -
3. BMI similarity 91.82 5.34 0.17* 0.36*** -
4. Perceived originality 3.56 1.44 0.38*** 0.14* 0.12 -
5. Age similarity 82.33 12.50 0.07 0.13 0.09 0.06 -

Note: Ethnicity: White = 140, Asian = 21, Afro-American = 14, Other = 27, Native = 1.
***p < .001; *p < .05.

F I G U R E 1 Path coefficients for Study

1—BMI similarity, perceived appearance
homophily, originality, and perceived
reliability of advertising model.
Standardized coefficients. Significance
level: ***p < .001.

(.128, ns). Hence, our findings indicate that perceived originality should appearance homophily in the context of thin advertising models. This
be excluded as an alternative explanation for the relationship between finding is important for marketing managers because recent advance-
BMI similarity and perceived reliability of the model. In addition, we ments in face recognition software facilitate a relatively accurate calcu-
were able to exclude similarity based on ethnicity or age as confounding lation of customers' BMI based on facial characteristics (Dantcheva
variables in our model because we included these variables as controls et al., 2018; Wen & Guo, 2013). In addition, Study 1 shows that BMI
in our model (ethnicity similarity: 0.021, ns; age similarity: 0.102, ns). similarity relates to how consumers rate the perceived reliability of the
Table 3 shows the regression coefficients and variance explained for advertising model. Thus, Study 1 provides initial evidence that BMI sim-
the different paths and stages of the model in more detail. ilarity is associated with managerially relevant outcome variables.
In addition, formal mediation tests (Table 4) based on boot-
strapping (Hayes, 2022) revealed that the mediating path from BMI
similarity to perceived reliability via perceived appearance homophily 4 | ST UDY 2
was statistically significant (.093, 95% CI 0.032 to 0.154). Furthermore,
the alternative mediation path through perceived originality was not Study 2 extends the initial findings from Study 1 in several important
statistically significant (.043, 95% CI 0.006 to 0.092). As shown in ways. First, it explores the relationship between BMI similarity for a thin,
Figure 1, the direct effect of BMI similarity on perceived reliability was female advertising model and women's perceived quality and purchase
not significant (.051, ns), which indicates full (rather than partial) media- intention for the apparel (a dress) endorsed by the advertising model (H3
tion (Pieters, 2017). Overall, the empirical results support our prediction and H4). To increase the external validity of our study, we emulated the
that perceived appearance homophily serves as a mediator between real conditions in which this phenomenon occurs by situating our study
similarity based on BMI and perceived reliability toward a human adver- in an online retail store and within a social media context. Furthermore,
tising model. This evidence suggests that in the context of thin adver- we included perceived vividness of the website as a moderator for the
tising models, women with BMIs similar to those of the advertising relationship between BMI similarity and perceived product quality (H6).
model perceive the advertising model to be more reliable.

4.1 | Method
3.3 | Discussion
4.1.1 | Participants and procedure
Study 1 demonstrates a positive relationship between BMI similarity
and appearance homophily for a human advertising model. Hence, the Three-hundred and thirty-one women (Mage = 38.8, SDage = 12.2)
results from Study 1 suggest that BMI similarity is a good proxy for were recruited through Amazon's MTurk to take part in an online
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 9

TABLE 3 Regression coefficients for Study 1.

Consequent

PAH (M1) PO (M2) PR (Y)

Antecedent Coef. SE z p Coef. SE z p Coef. SE z p

0
BMI (X) a1 0.373 0.058 6.415 <.001 a2 0.128 0.069 1.850 .064 c 0.051 0.069 0.743 .457
ES (C1) f11 0.101 0.046 2.202 .028 f11 0.113 0.052 2.203 .028 f11 0.021 0.069 0.301 .763
AS (C2) f21 0.038 0.05 0.754 .451 f21 0.043 0.056 0.761 .447 f21 0.102 0.058 1.755 .079
PAH (M1) b1 0.250 0.076 3.270 .001
PO (M2) b2 0.338 0.068 4.984 <.001
Constant iY 0 0.064 0 1 iY 0 0.069 0 1 iY 0 0.062 0 1
F(3,199) = 13.032, p < .001 F(3,199) = 1.488, p = .219 F(5.197) = 10.883, p < .001
R2 = .164 R2 = .022 R2 = .216

Note: Coefficients are standardized. Controls: ES = ethnic similarity, AS = age similarity.

Abbreviations: BMI, similarity based on body mass index; C1, first control; C2, second control; M1, first mediator; M2, second mediator; PAH, perceived
appearance homophily; PO, perceived originality; PR, perceived reliability; X, independent variable.

T A B L E 4 Mediation tests based on

Paths Coefficient Boot. SE Confidence intervala Equations
bootstrapping (Study 1).
BMI ! PAH ! PR 0.093 0.031 (0.032; 0.154) (a1b1)
BMI ! PO ! PR 0.043 0.025 (0.006; 0.092) (a2b2)
Direct effect BMI ! PR 0.051 0.069 (0.084, 0.187) (c0 )
Total effects 0.188 0.068 (0.054; 0.321) (c0 + a1b1 + a2b2)

Note: All coefficients are standardized.

Abbreviations: BMI, similarity based on body mass index; PAH, perceived appearance homophily;
PO, perceived originality; PR, perceived reliability.
a
10,000 bootstrap samples for bias-corrected confidence intervals at 95%.

study in exchange for a small cash incentive. The majority of the par- 4.1.3 | Moderating variable
ticipants had an annual household income of more than $50,000
(59.6% of the sample) and a minimum of an associate degree (67.8% Our conceptualization of website vividness follows a more compre-
of the sample). Participants entered a faux online store (see hensive approach in which a communication message includes both
Appendix B, Figures B1 and B2) where they were exposed to the central and peripheral elements. For example, Fortin and Dhola-
human advertising model (21 years of age, height 50 200 , 129.2 pounds, kia (2005, p. 395) indicate that they provided “enhanced vividness of
BMI of 20.5). Next, participants responded to questions regarding per- the message by means of colors, graphics, and animation” in their
ceived product quality and purchase intention for the dress endorsed research on social presence and involvement for web-based adver-
by the advertising model. tisements. Congruent with this conceptualization, we manipulated
website vividness in Study 2 by including animation of the top banner,
an animated cartoon image with sound of high-heeled steps, a
4.1.2 | Measures dynamic calendar, and an action button for the high vividness condi-
tion (compare Appendix B, Figures B1 and B2). The high and low viv-
The dependent variable—purchase intention for the dress endorsed idness groups resulted equivalent in terms of age, education level,
by the advertising model—was measured with a three-item, seven- income level, and time to complete the survey. As a manipulation
point Likert scale based on Wang et al. (2011). Perceived product check, participants rated perceived vividness on a four-item Likert
quality was measured with five items adapted from Jo's (2007) scale adapted from Coyle and Thorson (2001).
Absolute Quality of the Product scale. This scale measures con- In addition, we included the same control variables (age and eth-
sumers' evaluations of product quality without making comparisons nic similarity) as in Study 1. Appendix B, Table B1, shows the mea-
to any other product. The independent variable—BMI similarity— surement items and reliability assessment for Study 2. Following the
was calculated following the same procedure as in Study same criteria as in Study 1, validity and reliability of our measurement
1 (Appendix B, Figure B3 shows a distribution of BMI of each scales were satisfactory. The descriptive statistics and correlation
participant). matrix for Study 2 are shown in Table 5.
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
10 VAZQUEZ ET AL.

T A B L E 5 Descriptive statistics and

Variable M SD 1. 2. 3. 4. 5.
correlations for Study 2.
1. Purchase intention 3.33 1.71 -
2. Perceived product quality 4.26 1.17 0.73*** -
3. BMI similarity 94.16 4.87 0.24*** 0.17** -
4. Perceived vividness 3.93 1.67 0.34*** 0.28*** 0.24*** -
5. Age similarity 82.18 12.11 0.09 0.02 0.20*** 0.17** -

Note: Ethnicity: White = 254, Asian = 30, Afro-American = 19, Other = 23, Native = 5.
***p < .001; **p < .01.

4.2 | Results 4.3 | Discussion

4.2.1 | Manipulation check Study 2 provides evidence for our expectation that BMI similarity
is associated with perceived quality and purchase intention for
An ANOVA on the manipulation check question revealed that partici- fashion apparel. Hence, Study 2 shows that associations of BMI
pants in the high vividness condition rated perceived vividness of the similarity are not limited to perceptions of the advertising model.
website higher (M = 4.76, SD = 1.33) than those in the low vividness Rather, our findings suggest that this relationship relates to
condition (M = 2.93, SD = 1.48; F(1,329) = 139.8, p < .001). Hence, marketing-relevant outcome variables that ultimately relate to firm
our manipulation of website vividness was successful. performance and success. Congruent with prior research on vivid-
To test H3, H4, and H6, we used a moderated mediation model ness in communication messages (Frey & Eagly, 1993; Guadagno
which evaluates the effect of BMI similarity on purchase intention for et al., 2011), the positive association between BMI similarity and
the endorsed dress through perceived product quality. Perceived viv- perceived product quality diminishes when the retailer's website
idness moderates the relationship between BMI similarity and per- makes use of distracting elements and features aimed at increasing
ceived product quality. Hence, this model is equivalent to model 8 in the website's vividness. However, one can indeed also argue that
Hayes (2022). As in Study 1, all analyses were conducted using the this attenuating effect of vividness is a positive marketing outcome
processR package of the statistical software R. Figure 2 shows for those consumers whose BMI similarity with the advertising
the standardized path coefficients for this model. model is low. That is, when vividness is high, it does matter less
As predicted by H3, BMI similarity was positively related to per- whether BMI similarity is high or low for perceived product quality
ceived product quality (0.158, p < .01), which in turn had a positive of apparel promoted by the advertising model. In summary, the
association with purchase intention for the dress endorsed by the results from Study 2 provide process evidence and boundary con-
advertising model (0.706, p < .001) (H4). Furthermore, vividness (high ditions regarding how BMI similarity relates with purchase inten-
vs. low) had a negative moderating effect on the relationship between tion for fashion products in a context where consumers are
BMI similarity and purchase intention (0.080, p < .05). Hence, in exposed to advertising fashion models.
support of H6, perceived vividness acted as a suppressor for the rela-
tionship between BMI similarity and perceived product quality. Below
Table 6 presents the regression coefficients and Table 7 shows the 5 | ST UDY 3
mediation tests.
In Study 1, we showed the association between BMI similarity and
perceived appearance homophily, and in Study 2, we demonstrated
4.2.2 | Robustness check for BMI similarity that BMI similarity is positively related to perceived product quality
and purchase intention. However, the mechanism through which BMI
As a robustness check, we tested H3, H4, and H6 using perceived similarity affects perceived product quality may be driven also by
appearance homophily rather than BMI similarity as our independent other variables, apart from appearance homophily. For example, the
variable. The results of this model were consistent with the findings source credibility model (Ohanian, 1990; Wallace et al., 2021) sug-
presented here (see Web Appendix A, Figure I, Tables I and II). gests that trustworthiness and expertise influence attitudes toward
the source, whereas the source attractiveness model (McGuire, 1985;
Nguyen et al., 2013) suggests that familiarity with and likeability of
4.2.3 | Supplementary study on group similarity the source are related to attitudes. To test for these alternative expla-
nations, we ran five simple mediation models (Hayes PROCESS model
A supplementary study using a measure of group similarity (see Web 4) with BMI similarity as the IV and perceived product quality as the
Appendix B) provides additional evidence for the importance of simi- DV, and in each of the models we tested one of the five mediators:
larity measures for perceived product quality in the context of an appearance homophily, trustworthiness, expertise, familiarity, or
existing fashion clothing company (boohoo.com). likability.
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 11

F I G U R E 2 Path coefficients for Study

3—Mediating effect of perceived product
quality and moderating effect of website
vividness. Standardized coefficients.
Significance level: ***p < .001;
**p < .01; *p < .05.

TABLE 6 Regression coefficients for Study 2.

Consequent

PQ (M) PI (Y)

Antecedent Coef. SE z p Coef. SE z p

BMI (X) a1 0.158 0.050 3.145 .002 c10 0.093 0.040 2.347 .019
VIV (W) a2 0.018 0.054 0.340 .734 c20 0.020 0.037 0.540 .589
BMI:IN (X:W) a3 0.059 0.051 1.171 .243 c30 0.072 0.038 1.923 .055
ES (C1) f1 0.105 0.061 1.721 .089 g1 0.122 0.035 3.465 .001
AS (C2) f2 0.063 0.118 0.531 .595 g2 0.060 0.040 1.509 .131
PQ (M) b 0.710 0.025 28.846 <.001
F(5,325) = 3.080, p = .010 F(6,324) = 73.666, p < .001
R2 = .045 R2 = .577

Note: Coefficients are standardized. Controls: ES = ethnic similarity, AS = age similarity.

Abbreviations: BMI, similarity based on body mass index; C1, first control; C2, second control; M, mediator; PI, purchase intention; PQ, perceived product
quality; VIV, vividness; W, moderator; X, independent variable.

5.1 | Method similarity—was calculated following the same procedure as in Study

1 and Study 2 (Appendix C, Figure C2 shows a distribution of BMI of
5.1.1 | Participants and procedure each participant). Perceived appearance homophily was measured as
in Study 1 with an adapted scale from McCroskey et al. (1975), and
Two-hundred and eight women (Mage = 31.5, SDage = 8.9) were perceived quality was measured as in Study 2 with an adapted scale
recruited through Amazon's MTurk to take part in an online study in from Jo (2007). In Study 3, we used Ohanian's (1990) scale to measure
exchange for a small cash incentive. Participants entered a faux online trustworthiness of the source with five items and expertise of the
store (see Appendix B, Figure C1) where they were exposed to the source with four items. Familiarity with the source was measured with
human advertising model (21 years of age, height 50 700 , 161 pounds, three items from Zhou et al. (2010). Finally, likeability of the source
BMI of 25.2). All participants were exposed to the same advertising was measured with two items from Reinhard and Messner (2009). We
model. again included the same control variables (age and ethnic similarity) as
in Study 1 and Study 2. Appendix C, Table C1, shows the measure-
ment items and reliability assessment for Study 3. Following the same
5.1.2 | Measures criteria as in Study 1 and Study 2, validity and reliability of our mea-
surement scales were satisfactory. In addition, Study 3 was consistent
The dependent variable—perceived product quality—was measured with Study 1 and Study 2, as all analyses were conducted using the
with the same five items adapted from Jo's (2007) “Absolute Quality processR package of the statistical software R. The descriptive statis-
of the Product” scale as in Study 2. The independent variable—BMI tics and correlation matrix for Study 3 are shown in Table 8.
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
12 VAZQUEZ ET AL.

TABLE 7 Moderated mediation tests based on bootstrapping (Study 2) to explore conditional effects.

Mediation path: BMI ! PQ ! PI conditioned by VIV Coefficient Boot. SE Confidence intervala Equations
Indirect effect  VIV = low 0.074 0.050 (0.025; 0.172) (a1 + a3WLow)b
Indirect effect  VIV = high 0.158 0.053 (0.055; 0.261) (a1 + a3WHigh)b
Total indirect effect 0.116 0.037 (0.043; 0.188) (a1 + a3W)b
Direct effect 0.086 0.040 (0.009; 0.164) c1 0 + c3 0 W
Total effects  VIV = low 0.233 0.074 (0.087; 0.378) c10 + c30 WLow + (a1 + a3WLow)b
Total effects  VIV = high 0.172 0.069 (0.036; 0.308) c10 + c30 WHigh + (a1 + a3WHigh)b

Note: All coefficients are standardized.

Abbreviations: BMI, similarity based on body mass index; PI, purchase intention; PQ, perceived product quality; VIV, vividness.
a
10,000 bootstrap samples for bias-corrected confidence intervals at 95%.

TABLE 8 Descriptive statistics and correlations for Study 3.

Variable M SD 1. 2. 3. 4. 5. 6. 7. 8.
1. Perceived product quality 5.80 0.69 -
2. Source trustworthiness 5.92 0.73 0.63*** -
3. Source expertise 5.95 0.72 0.58*** 0.83*** -
4. Source likeability 6.02 0.79 0.62*** 0.76*** 0.75*** -
5. Source familiarity 5.72 0.96 0.39*** 0.56*** 0.62*** 0.47*** -
6. Perceived appearance homophily 5.79 0.85 0.36*** 0.57*** 0.52*** 0.51*** 0.39*** -
7. BMI similarity 96.19 2.52 0.04 0.14** 0.08 0.14 0.08 0.19** -

8. Age similarity 92.24 8.61 0.02 0.13 0.18** 0.12 0.24*** 0.17* .10 -

Note: Ethnicity: White = 195, Asian = 1, Other = 6, Native = 3.

***p < .001; **p < .01; *p < .05.

5.2 | Results by the fact that the mediators in Study 3 are naturally correlated
among each other (see also the correlation matrix for Study 3 in
As predicted, the indirect effect of perceived appearance homophily Table 8). Hence, our interpretation focuses on the results from the
was positive and statistically significant (.069, 95% CI 0.016 to 0.123) simple mediation models.
in the simple mediation model. Furthermore, the indirect effects of
trustworthiness (0.091, 95% CI 0.007 to 0.176) and likeability (0.089,
95% CI 0.004 to 0.174) were positive and statistically significant. 5.3 | Discussion
However, the indirect effects of expertise (0.048, 95% CI 0.032 to
0.128) and familiarity (0.030, 95% CI 0.028 to 0.087) were not sta- The results from Study 3 suggest that BMI similarity is positively
tistically significant since the confidence interval crosses zero. related to perceived product quality, and that this relationship is medi-
Figure 3 shows the simple mediation models for the three significant ated by perceived appearance homophily, thereby supporting our
mediation paths through appearance homophily, trustworthiness, and main expectations regarding the role of BMI similarity and appearance
likeability. homophily. It is important to note that BMI similarity is an objective
Following Stoner et al. (2023), we also ran a parallel mediation measure based on consumers' weight and height, and the causality of
model with all mediators included in the same model (Figure 4). For the relationships in Study 3 (as well as in Study 1 and 2) must there-
this parallel mediation model with five mediators, none of the indirect fore flow from BMI similarity to any of the latent measures in our
effects were significant (appearance homophily: 0.003, 95% CI models. In other words, it is not conceivable that consumers' weight
0.031 to 0.025; trustworthiness: 0.052, 95% CI 0.018 to 0.123; and height could change based on perceptions of, for example,
likeability: 0.052, 95% CI 0.009 to 0.113; expertise: 0.008, 95% CI appearance homophily, perceived product quality, or purchase inten-
0.018 to 0.034; familiarity: 0.004, 95% CI 0.010 to 0.018). A highly tion for the apparel endorsed by the advertising model. We believe
plausible explanation for this finding is that the total effect in this par- that this characteristic of BMI similarity as a directly observable mea-
allel mediation model is decomposed into a direct effect and five indi- sure (similar to other directly observable measures such as age, gen-
rect effects with equal signs, with the result that each indirect effect der, or income) makes it a powerful variable for marketing theory and
becomes relatively small (compare, e.g., Pieters, 2017, for the decom- practice. At the same time, Study 3 reveals (not surprisingly) that
position of total effects in mediation models). This issue is aggravated other variables than appearance homophily can explain the effect of
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 13

F I G U R E 3 Path coefficients for Study

3—Simple mediation models for
(a) appearance homophily,
(b) trustworthiness, and (c) likeability.
Standardized coefficients. Significance
level: ***p < .001; *p < .05.

BMI similarity on perceived product quality. Specifically, Study 3 finds insights on how BMI similarity relates to perceived appearance homo-
that perceived trustworthiness and likeability are also influenced by phily and marketing-relevant outcome variables for apparel retail prod-
BMI similarity. Hence, even though Study 3 finds support for the role ucts. Specifically, the results demonstrate that both perceived
of appearance homophily as a mechanism explaining the relationship appearance homophily and BMI similarity can be associated with the
between BMI similarity and perceived product quality, alternative perceived reliability of advertising models (Study 1) and perceived prod-
explanations exist in terms of trustworthiness and likeability. uct quality perceptions (Study 2). Furthermore, a series of mediation
models shows that in addition to perceived appearance homophily,
trustworthiness and likeability are identified as plausible alternative
6 | G E N E R A L D I S C U S SI O N explanations for the relationship between BMI similarity and perceived
product quality (Study 3). Vividness of the retailer's website is identified
Drawing on the results from three empirical studies and a supplemen- as an important boundary condition for these effects which attenuates
tary study2 on group similarity reported in the Web Appendix, the cur- the relationship between BMI similarity and perceived product quality.
rent research contributes to marketing theory and practice by providing

2
6.1 | Theoretical implications
All studies used samples from Amazon's MTurk, which have been shown to be of better
quality than those from professional panels (Kees et al., 2017). Additionally, we followed
steps recommended by Aguinis et al. (2021) in combination with time controls and click The results of the current research yield important contributions to
counts from Qualtrics (e.g., Vazquez et al., 2023) to increase the quality of the samples. Some
of these steps included removing responses with no input activity or IP address, as these
psychology and marketing theory by providing an alternative theoreti-
were suspected web-robots; as well as removing responses of inattentive participants who cal view from that of previous research about female portrayals
did not select their gender as female when the questionnaire instructions invited only female
involving the body size (e.g., Åkestam et al., 2017; Janssen &
participants; or the lack of attention to correctly filling in the random unique number
generated in each questionnaire to validate it. Paas, 2014). First, our findings contribute to the extant literature on
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
14 VAZQUEZ ET AL.

F I G U R E 4 Path coefficients for Study

3—Parallel mediation model for the
association between BMI similarity and
perceived product quality. Standardized
coefficients. Significance level: **p < .01;
*p < .05;  p < .10.

homophily (Bilgicer et al., 2015; McPherson et al., 2001) and similarity draws cognitive resources, information processing will be hindered.
(Rodrigues et al., 2017; Thakor et al., 2008) by showing the link Even though previous research predominantly reports situations in
between traditional appearance homophily (Ladhari et al., 2020; which vivid messages have positive outcomes on attitudes toward the
McCroskey et al., 2006) and our refined construct of BMI similarity. message and message engagement (Coyle & Thorson, 2001; Fortin &
Specifically, the results from Study 1 demonstrate that BMI similarity Dholakia, 2005; Yousaf et al., 2020), our research shows that vivid-
is a parsimonious measure that captures major aspects of appearance ness can attenuate the processing of key information (such as differ-
homophily without the need to gather additional customer informa- ences in BMI between the advertising model and the customer).
tion, such as body shape, skin color, or facial features. Furthermore, Hence, in line with Frey and Eagly (1993), we find that website vivid-
our research demonstrates that consumers' BMI similarity to a human ness can distract the viewer and undermine information processing.
advertising model is positively associated with marketing-relevant This implies that vividness in communication messages can act as a
downstream variables, such as the perceived reliability of the advertis- double-edged sword: For consumers with high BMI similarity to the
ing model, perceived product quality, and purchase intention. advertising model, website vividness decreases the relationship
Second, our research reconciles conflicting streams of literature on between BMI similarity and perceived quality of the apparel promoted
the BMI of advertising models which either predict higher advertising by the advertising model, which is indeed an undesired outcome for
effectiveness for relatively thin (Borland & Akram, 2007; Janssen & marketers. On the other hand, website vividness has the effect that
Paas, 2014) or average-size models (Lou et al., 2019). Specifically, our BMI similarity matters less for perceived product quality, which is a
research finds that rather than an ideal BMI value fitting all consumers positive marketing outcome for those consumers who score very low
simultaneously, it is the match between the consumer's BMI with the on BMI similarity. Finally, the present work contributes to theory of
BMI of the advertising model that covaries with marketing-relevant social comparison processes (e.g., Festinger, 1954; Mussweiler, 2003)
outcome variables in the context of a thin, female advertising model. by showing that perceptions of BMI depend on the BMI of the
Furthermore, our research extends those streams of BMI literature that observer through an implicit comparison of an advertising model's
focus predominantly on obesity and health-related public policy impli- BMI with one's own BMI.
cations (Cornil et al., 2022; Zlatevska et al., 2018).
Third, the current research adds to the extant literature on vivid-
ness in communication by demonstrating that vivid elements 6.2 | Managerial implications
(e.g., peripheral animations on a website) can inhibit the processing of
relevant information. Keller and Block (1997) suggested that when a The results from our research also have important managerial implica-
vivid communication message is less resource demanding, the vivid- tions. First, building on our finding that BMI similarity (rather than an
ness of the message is going to facilitate the processing of the core absolute BMI value) positively relates to perceived product quality
content of the message. However, when the vividness of the message and purchase intention for female consumers, dynamic displays of
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 15

advertising models that adjust to the consumer's own BMI can pro- holds in the context of plus-size advertising models, given that some
vide a competitive advantage as compared to traditional marketing marketers have begun to employ plus-size advertising models in their
strategies relying on static reproductions of apparel models. Apparel marketing campaigns (Cinelli & Yang, 2016; Joo & Wu, 2021;
retailers aiming at enhancing their marketing effectiveness could Pounders, 2018). For example, future research on BMI similarity might
implement BMI recognition technology (Wen & Guo, 2013) in their build on the research design from McFerran et al. (2010a, 2010b) who
online and physical stores, which would allow them to dynamically tai- manipulated body type by exposing participants to either thin or
lor displays of apparel models according to their customers' BMI. Such heavy people around them. Such an extension of our research would
an approach would also help marketers with an organic inclusion of be timely and relevant since scholars have raised concerns about how
plus-size models in their marketing campaigns (Joo & Wu, 2021). This to mitigate the negative emotional effects of promoting thinness and
is consistent with the benefits from message personalization, which idealized images of “perfect” thin models to young women (Holland &
has proved to be effective in digital environments to target diverse Tiggemann, 2016; Kim & Chock, 2015; Perloff, 2014).
retail consumers as in the case of social commerce (e.g., Vazquez Third, BMI similarity is a relatively stable individual difference
et al., 2023). Beyond the online context, marketers could employ characteristic, similar to, for example, age or marital status. It can
smart shelves with dynamic LED screens displaying marketing mes- indeed change over time, but it is not influenced by situational factors.
sages based on the proximity and BMI of shoppers (Grewal Hence, we argue that BMI similarity can influence consumer judg-
et al., 2021). Because personalizing advertising based on BMI similar- ments and perceptions, such as perceived homophily, liking of the
ity promotes body diversity and inclusivity among apparel models and advertising model, or perceived product quality, but none of these
consumers, the use of BMI similarity could be crucial for brands that consumer perceptions can situationally change BMI similarity, in the
appeal for societal trends such as diversity and inclusion of varied same vein that situationally triggered consumer perceptions cannot
body types or body builds (e.g., heavy body type or overweight); change the age or marital status of a respondent. Nevertheless, we
which is in line with social movements promoting body positivity, fat acknowledge that our work builds on associations rather than causal
acceptance, plus-size, health at every size, anti-diet, and body neutral- relationships triggered by the manipulation of variables. Hence, even
ity. Furthermore, our research suggests that marketing managers aim- though BMI cannot be directly manipulated since it is based on the
ing at making use of the effects of BMI similarity should be careful height and weight of respondents, future research could manipulate
with deploying overly vivid websites in their customer communication perceived appearance homophily to show causality on outcome vari-
to avoid distracting consumers from the core message. ables such as perceived reliability of the advertising model, perceived
product quality, and purchase intention. Furthermore, future research
could follow very recent recommendations for more stringent tests of
7 | L I M I T A TI O NS A ND F U T U R E R E S E A R CH causality in models that measure both mediators and outcome vari-
ables (Laghaie & Otter, 2023).
Even though BMI has been widely accepted among health experts, it is Finally, the current research investigates the relationship between
an imperfect measure of body appearance. For example, BMI typically BMI similarity and website vividness in the context of online fashion
fails to detect the substitution of lean with fat body tissue that accom- apparel. We argue that this specific context is meaningful because
panies normal aging (Prentice & Jebb, 2001). Furthermore, an individual fashion apparel is closely related to the body and may in fact be seen
engaging in exercising or body building may have the same weight and as an extension of the self (Ahuvia, 2005). However, bodily appear-
height as an overweight person, but body appearance in such a case ance of advertising models may also be relevant for product catego-
may be quite dissimilar. Hence, BMI as a measure of appearance simi- ries that are not directly related to the body. Thus, future research
larity toward the advertising model may be biased, which in turn would may extend our work to product categories such as food, consumer
inflate or deflate relationships with downstream marketing variables electronics, automobiles, or services.
such as product perceptions and purchase intention for fashion apparel.
Future research may try to collect additional data on respondents' body ACKNOWLEDG MENTS
fat to determine potential bias in the BMI similarity measure. The authors are grateful for the helpful comments and guidance by
Second, since the female thin-ideal still prevails in mainstream two anonymous reviewers and the editors during the review process.
advertising (Cohen et al., 2017; Volonté, 2019), our research inten- The first author would also like to thank the National System of
tionally focuses on advertising models with a comparably low BMI Researchers (SNI) of CONAHCYT Mexico for providing incentives to
(Study 1: 17.8; Study 2: 20.5) but also extends these findings to a conduct this academic research.
slightly overweight advertising model (Study 3: 25.2). Even though
these BMI values are not necessarily exceptional and fall in the range CONFLIC T OF INTER E ST STATEMENT
of healthy weight (Centers for Disease Control and Prevention, 2021), The authors state there are no conflicts of interest.
they are substantially below the average US female population of
about 28.42 (Centers for Disease Control and Prevention, 2020). DATA AVAILABILITY STAT EMEN T
Thus, there is an opportunity for future research to investigate The data that support the findings of this study are available from the
whether the effect of BMI similarity shown in the current research corresponding author upon reasonable request.
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
16 VAZQUEZ ET AL.

ORCID Coyle, J. R., & Thorson, E. (2001). The effects of progressive levels of inter-
Erik Ernesto Vazquez https://orcid.org/0000-0002-7592-1571 activity and vividness in web marketing sites. Journal of Advertising,
30(3), 65–77.
Reto Felix https://orcid.org/0000-0002-3864-2397
D'Alessandro, S., & Bill Chitty, B. (2011). Real or relevant beauty? Body
shape and endorser effects on brand attitude and body image. Psychol-
RE FE R ENC E S ogy & Marketing, 28(8), 843–878.
Aguinis, H., Villamor, I., & Ramani, R. S. (2021). MTurk research: Review Dantcheva, A., Bremond, F., & Bilinski, P. (2018). Show me your face and I
and recommendations. Journal of Management, 47(4), 823–837. will tell you your height, weight and body mass index. Proceedings
Ahuvia, A. C. (2005). Beyond the extended self: Loved objects and con- of the International Conference on Pattern Recognition, 3555–3560.
sumers' identity narratives. Journal of Consumer Research, 32(1), Beijing, China.
171–184. Das Guru, R. R., & Paulssen, M. (2020). Customers' experienced product
Åkestam, N., Rosengren, S., & Dahlen, M. (2017). Advertising “like a girl”: quality: Scale development and validation. European Journal of Market-
Toward a better understanding of “femvertising” and its effects. Psy- ing, 54(4), 645–670.
chology and Marketing, 34(8), 795–806. De Bruyn, A., & Lilien, G. L. (2008). A multi-stage model of word-of-mouth
Bagozzi, R. P., & Khoshnevis, M. (2022). How and when brand coolness influence through viral marketing. International Journal of Research in
transforms product quality judgments into positive word of mouth and Marketing, 25(3), 151–163.
intentions to buy/use. Journal of Marketing Theory and Practice, 31, Dwyer, S., Richard, O., & Shepherd, C. D. (1998). An exploratory study of
383–402. https://doi.org/10.1080/10696679.2022.2081925 gender and age matching in the salesperson-prospective customer
Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation dyad: Testing similarity-performance predictions. Journal of Personal
models. Journal of the Academy of Marketing Science, 16, 74–94. Selling and Sales Management, 18(4), 55–69.
Balabanis, G., & Chatzopoulou, E. (2019). Under the influence of a blogger: Ertug, G., Brennecke, J., Kovács, B., & Zou, T. (2022). What does homo-
The role of information-seeking goals and issue involvement. Psychol- phily do? A review of the consequences of homophily. Academy of
ogy & Marketing, 36(4), 342–353. Management Annals, 16(1), 38–69.
Barry, B. (2014). Selling whose dream? A taxonomy of aspiration in fashion Felix, R., & Garza, M. R. (2012). Rethinking worldly possessions: The rela-
imagery. Fashion, Style & Popular Culture, 1(2), 175–192. tionship between materialism and body appearance for female con-
Bekk, M., Spörrle, M., Völckner, F., Spieß, E., & Woschée, R. (2017). What sumers in an emerging economy. Psychology & Marketing, 29(12),
is not beautiful should match: How attractiveness similarity affects 980–994.
consumer responses to advertising. Marketing Letters, 28, 509–522. Festinger, L. (1954). A theory of social comparison processes. Human Rela-
Bilgicer, T., Jedidi, K., Lehmann, D. R., & Neslin, S. A. (2015). Social conta- tions, 7(2), 117–140.
gion and customer adoption of new sales channels. Journal of Retailing, Filieri, R., Acikgoz, F., Li, C., & Alguezaui, S. (2023). Influencers' “organic”
91(2), 254–271. persuasion through electronic word of mouth: A case of sincerity over
Bjornsdottir, R. T., Hehman, E., Agboh, D., & Rule, N. O. (2021). Parsing brains and beauty. Psychology & Marketing, 40(2), 347–364.
the mechanisms underlying ingroup facial resemblance. Personality and Fornell, C., & Larcker, D. F. (1981). Structural equation models with unob-
Social Psychology Bulletin, 48(5), 782–792. servable variables and measurement error: Algebra and statistics. Jour-
Borau, S., & Bonnefon, J.-F. (2017). The advertising performance of nal of Marketing Research, 18(3), 382–388.
non-ideal female models as a function of viewers' body mass index: A Fortin, D. R., & Dholakia, R. R. (2005). Interactivity and vividness effects
moderated mediation analysis of two competing affective pathways. on social presence and involvement with a web-based advertisement.
International Journal of Advertising, 36(3), 457–476. Journal of Business Research, 58(3), 387–396.
Borgatti, S. P., & Halgin, D. S. (2011). On network theory. Organization Sci- Fratto, G. M., Jones, M. R., & Cassill, N. L. (2006). An investigation of com-
ence, 22(5), 1168–1181. petitive pricing among apparel retailers and brands. Journal of Fashion
Borland, H., & Akram, S. (2007). Age is no barrier to wanting to look good: Marketing and Management: An International Journal, 10(4), 387–404.
Women on body image, age and advertising. Qualitative Market Frey, K. P., & Eagly, A. H. (1993). Vividness can undermine the persuasive-
Research: An International Journal, 10(3), 310–333. ness of messages. Journal of Personality and Social Psychology, 65(1),
Byrne, D. (1971). The attraction paradigm. Academic Press. 32–44.
Campbell, N. C. G., Graham, J. L., Jolibert, A., & Gunther, H. (1988). Mar- Fu, S., Xu, Y., & Yan, Q. (2019). Enhancing the parasocial interaction rela-
keting negotiations in France, Germany, the United Kingdom, and the tionship between consumers through similarity effects in the context
United States. Journal of Marketing, 52(2), 49–62. of social commerce. Journal of Strategic Marketing, 27(2), 100–118.
Casalo
, L. V., Flavián, C., & Ibáñez-Sánchez, S. (2020). Influencers on Insta- Golder, P. N., Mitra, D., & Moorman, C. (2012). What is quality? An integra-
gram: Antecedents and consequences of opinion leadership. Journal of tive framework of processes and states. Journal of Marketing, 76(4), 1–23.
Business Research, 117, 510–519. Grewal, D., Gauri, D. K., Roggeveen, A. L., & Sethuraman, R. (2021). Strate-
Centers for Disease Control and Prevention. (2020). National health and gizing retailing in the new technology area. Journal of Retailing, 97(1),
nutrition examination survey, 2017–2018. https://wwwn.cdc.gov/ 6–12.
Nchs/Nhanes/2017-2018/BMX_J.htm Guadagno, R. E., Rhoads, K. V. L., & Sagarin, B. J. (2011). Figural vividness
Centers for Disease Control and Prevention. (2021). Defining adult over- and persuasion: Capturing the “elusive” vividness effect. Personality
weight & obesity. https://www.cdc.gov/obesity/adult/defining.html and Social Psychology Bulletin, 37(5), 626–638.
Cinelli, M. D., & Yang, L. (2016). The role of implicit theories in evaluations Hayes, A. F. (2022). Introduction to mediation, moderation, and conditional
of “plus-size” advertising. Journal of Advertising, 45(4), 472–481. process analysis, second edition: A regression-based approach (3rd ed.).
Cohen, R., Newton-John, T., & Slater, A. (2017). The relationship between The Guilford Press.
Facebook and Instagram appearance-focused activities and body image Heiser, R. S., Sierra, J. J., & Torres, I. M. (2008). Creativity via cartoon
concerns in young women. Body Image, 23, 183–187. spokespeople in print ads capitalizing on the distinctiveness effect.
Cornil, Y., Plassmann, H., Aron-Wisnewsky, J., Poitou-Bernert, C., Journal of Advertising, 37(4), 75–84.
Clément, K., Chabert, M., & Chandon, P. (2022). Obesity and respon- Hildebrand, D., Rubin, D., Hadi, R., & Kramer, T. (2021). Flavor fatigue:
siveness to food marketing before and after bariatric surgery. Journal Cognitive depletion influences consumer enjoyment of complex fla-
of Consumer Psychology, 32(1), 57–68. vors. Journal of Consumer Psychology, 31(1), 103–111.
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 17

Holland, G., & Tiggemann, M. (2016). A systematic review of the impact of McCroskey, J. C., Richmond, V. P., & Daly, J. A. (1975). The development
the use of social networking sites on body image and disordered eat- of a measure of perceived homophily. Human Communication Research,
ing outcomes. Body Image, 17, 100–110. 1(4), 323–332.
Homburg, C., & Stock, R. M. (2005). Exploring the conditions under which McCroskey, L. L., McCroskey, J. C., & Richmond, V. P. (2006). Analysis and
salesperson work satisfaction can lead to customer satisfaction. Psy- improvement of the measurement of interpersonal attraction and
chology & Marketing, 22(5), 393–420. homophily. Communication Quarterly, 54(1), 1–31.
Hudders, L., & De Jans, S. (2022). Gender effects in influencer marketing: McFerran, B., Dahl, D. W., Fitzsimons, G. J., & Morales, A. C. (2010a). I'll
An experimental study on the efficacy of endorsements by same- have what she's having: Effects of social influence and body type on
vs. other-gender social media influencers on Instagram. International the food choices of others. Journal of Consumer Research, 36(6),
Journal of Advertising, 41(1), 128–149. 915–929.
Ismagilova, E., Slade, E., Ranac, N. P., & Dwivedi, Y. K. (2020). The effect of McFerran, B., Dahl, D. W., Fitzsimons, G. J., & Morales, A. C. (2010b).
characteristics of source credibility on consumer behaviour: A meta- Might an overweight waitress make you eat more? How the body type
analysis. Journal of Retailing and Consumer Services, 53, 101736. of others is sufficient to alter our food consumption. Journal of Con-
Janssen, D. M., & Paas, L. J. (2014). Moderately thin advertising models sumer Psychology, 20(2), 146–151.
are optimal, most of the time: Moderating the quadratic effect of McGuire, W. J. (1985). Attitudes and attitude change. In G. Lindzey & E.
model body size on ad attitude by fashion leadership. Marketing Let- Aronson (Eds.), Handbook of social psychology (Vol. 2, pp. 233–346).
ters, 25, 167–177. Random House.
Jo, M. (2007). Should a quality sub-brand be located before or after the McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Beards of a feather:
parent brand? An application of composite concept theory. Journal of Homophily in social networks. Annual Review of Sociology, 27,
the Academy of Marketing Science, 35(2), 184–196. 415–444.
Joo, B. R., & Wu, J. (2021). The impact of inclusive fashion advertising with Montoya, R. M., & Horton, R. S. (2013). A meta-analytic investigation of
plus-size models on female consumers: The mediating role of brand the processes underlying the similarity-attraction effect. Journal
warmth. Journal of Global Fashion Marketing, 12(3), 260–273. of Social and Personal Relationships, 30(1), 64–94.
Kees, J., Berry, C., Burton, S., & Sheehan, K. (2017). An analysis of data Moon, K. W., & Hong, S. (2021). Package ‘processR’. R Package version
quality: Professional panels, student subject pools, and Amazon's 0.2.6. https://cran.r-project.org/web/packages/processR/processR.pdf
Mechanical Turk. Journal of Advertising, 46(1), 141–155. Mussweiler, T. (2003). Comparison processes in social judgment: Mecha-
Keller, P. A., & Block, L. G. (1997). Vividness effects: A resource-matching nisms and consequences. Psychological Review, 110(3), 472–489.
perspective. Journal of Consumer Research, 24(3), 295–304. Navarro, J., Peña, J., Cebolla, A., & Baños, R. (2022). Can avatar appearance
Kim, D. Y., & Kim, H.-Y. (2021). Trust me, trust me not: A nuanced view of influence physical activity? User-avatar similarity and proteus effects
influencer marketing on social media. Journal of Business Research, on cardiac frequency and step counts. Health Communication, 37(2),
134, 223–232. 222–229.
Kim, J. W., & Chock, T. M. (2015). Body image 2.0: Associations between Nguyen, B., Melewar, T. C., & Chen, J. (2013). A framework of brand like-
social grooming on Facebook and body image concerns. Computers in ability: An exploratory study of likeability in firm-level brands. Journal
Human Behavior, 48, 331–339. of Strategic Marketing, 21(4), 368–390.
Ladhari, R., Massa, E., & Skandrani, H. (2020). YouTube vloggers' popular- Ohanian, R. (1990). Construction and validation of a scale to measure
ity and influence: The roles of homophily, emotional attachment, and celebrity endorsers' perceived expertise, trustworthiness, and attrac-
expertise. Journal of Retailing and Consumer Services, 54, 102027. tiveness. Journal of Advertising, 19(3), 39–52.
Laghaie, A., & Otter, T. (2023). Measuring evidence for mediation in the O'Quin, K., & Besemer, S. P. (1989). The development, reliability, and valid-
presence of measurement error. Journal of Marketing Research, 60(5), ity of the revised creative product semantic scale. Creativity Research
847–869. https://doi.org/10.1177/00222437231151873 Journal, 2(4), 267–278.
Lazarsfeld, P. F., & Merton, R. K. (1954). Friendship as a social process: A sub- Pechman, C., Yoon, K. E., Trapido, D., & Prochaska, J. J. (2021). Perceived
stantive and methodological analysis. In M. Berger, T. Abel, & C. H. Page costs vs. actual benefits of demographic self-disclosure in online sup-
(Eds.), Freedom and control in modern society (pp. 18–66). Van Nostrand. port groups. Journal of Consumer Psychology, 31(3), 450–477.
Leonhardt, J. M., Pezzuti, T., & Namkoong, J.-E. (2020). We're not so dif- Perloff, R. M. (2014). Social media effects on young women's body image
ferent: Collectivism increases perceived homophily, trust, and seeking concerns: Theoretical perspectives and an agenda for research. Sex
user-generated product information. Journal of Business Research, 112, Roles, 71(11), 363–377.
160–169. Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Cen-
Li, Y., & Peng, Y. (2021). Influencer marketing: Purchase intention and its tral and peripheral routes to attitude change. Springer.
antecedents. Marketing Intelligence & Planning, 39(7), 960–978. Pieters, R. (2017). Meaningful mediation analysis: Plausible causal infer-
Lin, L., Hoegg, J., & Aquino, K. (2018). When beauty backfires: The effects ence and informative communication. Journal of Consumer Research,
of server attractiveness on consumer taste perceptions. Journal of 44(3), 692–716.
Retailing, 94(3), 296–311. Pounders, K. (2018). Are portrayals of female beauty in advertising finally
Lou, C., Tse, C. H., & Lwin, M. O. (2019). “Average-sized” models do sell, changing. Journal of Advertising Research, 58(2), 133–137.
but what about east asia? A cross-cultural investigation of U.S. and Prendergast, G., Ko, D., & Yuen, S. Y. V. (2010). Online word of mouth and
Singaporean women. Journal of Advertising, 48(5), 512–531. consumer purchase intentions. International Journal of Advertising,
Lv, X., Liang, Y., Luo, J., & Liu, Y. (2022). Icing on the cake or gilding the 29(5), 687–708.
lily? The impact of high-modified model images on purchase intention. Prentice, A. M., & Jebb, S. A. (2001). Beyond body mass index. Obesity
Journal of Retailing and Consumer Studies, 68, 103078. Reviews, 2(3), 141–147.
MacKenzie, S. B., Podsakoff, P. M., & Podsakoff, N. P. (2011). Construct Racherla, P., Mandviwalla, M., & Connolly, D. J. (2012). Factors affecting
measurement and validation procedures in MIS and behavioral consumers' trust in online product reviews. Journal of Consumer Behav-
research: Integrating new and existing techniques. MIS Quarterly, iour, 11(2), 94–104.
35(2), 293–334. Reinhard, M. A., & Messner, M. (2009). The effects of source likeability
McCracken, G. (1989). Who is the celebrity endorser? Cultural foundations of and need for cognition on advertising effectiveness under explicit per-
the endorsement process. Journal of Consumer Research, 16(3), 310–321. suasion. Journal of Consumer Behaviour, 8(4), 179–191.
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
18 VAZQUEZ ET AL.

Risselada, H., Verhoef, P. C., & Bijmolt, T. H. (2014). Dynamic effects of proposed remedies. Journal of the Academy of Marketing Science, 44,
social influence and direct marketing on the adoption of high-technol- 119–134.
ogy products. Journal of Marketing, 78(2), 52–68. Wallace, L. E., Simon, K. A., & Wegener, D. T. (2021). Lay concepts of
Rodrigues, D., Lopes, D., Alexopoulos, T., & Goldenberg, L. (2017). A source likeability, trustworthiness, expertise, and power: A prototype
new look at online attraction: Unilateral initial attraction and the analysis. Behavior Research Methods, 53, 1188–1201.
pivotal role of perceived similarity. Computers in Human Behavior, Wang, Y. J., Minor, M. S., & Wei, J. (2011). Aesthetics and the online shop-
74, 16–25. ping environment: Understanding consumer responses. Journal of
Semaan, R. W., Kocher, B., & Gould, S. (2018). How well will this brand Retailing, 87(1), 46–58.
work? The ironic impact of advertising disclosure of body-image Wen, L., & Guo, G. (2013). A computational approach to body mass index
retouching on brand attitudes. Psychology & Marketing, 35(10), prediction from face images. Image and Vision Computing, 31(5),
766–777. 392–400.
Soh, H., Reid, L. N., & King, K. W. (2009). Measuring trust in advertising: White, A., & Smith, B. L. (2001). Assessing advertising creativity using the
Development and validation of the ADTRUST scale. Journal of Adver- creative product semantic scale. Journal of Advertising Research, 41(6),
tising, 38(2), 83–103. 27–34.
Sokolova, K., & Kefi, H. (2020). Instagram and YouTube bloggers promote Yousaf, A., Amin, I., Jaziri, D., & Mishra, A. (2020). Effect of message
it, why should I buy? How credibility and parasocial interaction influ- orientation/vividness on consumer engagement for travel brands on
ence purchase intentions. Journal of Retailing and Consumer Services, social networking sites. Journal of Product & Brand Management, 30(1),
53, 101742. 44–57.
Sokolova, T., & Krishna, A. (2016). Take it or leave it: How choosing versus Zebal, M. A., & Jackson, F. H. (2019). Cues for shaping purchase of local
rejecting alternatives affects information processing. Journal of Con- retail apparel clothing brands in an emerging economy. International
sumer Research, 43(4), 614–635. Journal of Retail and Distribution Management, 47(10), 1013–1028.
Steuer, J. (1992). Defining virtual reality: Dimensions defining telepre- Zhou, L., Zhiyong, Y., & Hui, M. K. (2010). Non-local or local brands? A
sence. Journal of Communication, 42(4), 73–93. multi-level investigation into confidence in brand origin identification
Stoner, J. L., Felix, R., & Stadler Blank, A. (2023). Best practices for imple- and its strategic implications. Journal of Academy of Marketing Science,
menting experimental research methods. International Journal of Con- 38(2), 202–218.
sumer Studies, 47(4), 1579–1595. Zlatevska, N., Neumann, N., & Dubelaar, C. (2018). Mandatory calorie dis-
Stone-Romero, E. F., Stone, D. L., & Grewal, D. (1997). Development of a closure: A comprehensive analysis of its effect on consumers and
multidimensional measure of perceived product quality. Journal of retailers. Journal of Retailing, 94(1), 89–101.
Quality Management, 2(1), 87–111.
Streukens, S., & Andreassen, T. W. (2013). Customer preferences for front-
line employee traits: Homophily and heterophily effects. Psychology &
AUTHOR BIOGRAPHIES
Marketing, 30(12), 1043–1052.
Thakor, M. V., Suri, R., & Saleh, K. (2008). Effects of service setting and
other consumers' age on the service perception of young consumers. Erik Ernesto Vazquez is an Associate Professor of Digital Business
Journal of Retailing, 84(2), 137–149.
and Data Analytics and the Director of the Graduate Studies in
Torres, I. M. (2007). A tale of two theories: Sympathy or competition?
Journal of Business Research, 60(3), 197–205. Digital Business Analytics at EMLV Business School in Paris. He
Van der Land, S. F., Schouten, A. P., Feldberg, F., Huysman, M., & van den holds a Doctoral degree from the University of Newcastle upon
Hooff, B. (2015). Does avatar appearance matter? How team visual Tyne, UK and an MBA from the University of North Carolina,
similarity and member-avatar similarity influence virtual team perfor-
USA. Prof. Vazquez has led research projects that have been pub-
mance. Human Communication Research, 41(1), 128–153.
Van Esch, P., Arli, D., Castner, J., Talukdar, N., & Northey, G. (2018). Con- lished in journals such as Journal of Retailing and Consumer Ser-
sumer attitudes towards bloggers and paid blog advertisements: vices, Electronic Markets, and Journal of Research in Interactive
What's new? Marketing Intelligence & Planning, 36(7), 778–793. Marketing. He is also a national research member Level I of the
Vazquez, E. E. (2020). Effects of enduring involvement and perceived con-
CONAHCYT (also known as CONACYT), a member of the edito-
tent vividness on digital engagement. Journal of Research in Interactive
rial board of the Anahuac Journal, and an executive advisor to
Marketing, 14(1), 1–16.
Vazquez, E. E. (2021). Effect of an e-retailer's product category and social international companies, scaleups, and startups.
media platform selection on perceived quality of e-retail products.
Electronic Markets, 31, 139–157. Reto Felix is an Associate Professor of Marketing at the Univer-
Vazquez, E. E., Patel, C., Alvidrez, S., & Siliceo, L. (2023). Images, reviews, sity of Texas Rio Grande Valley, USA and holds a PhD from the
and purchase intention on social commerce: The role of mental imag- University of St. Gallen, Switzerland. His research focuses on
ery vividness, cognitive and affective social presence. Journal of Retail-
green consumption, consumer well-being, and new technologies.
ing and Consumer Services, 74, 103415.
Venkatesh, A., Joy, A., Sherry, J. F., & Deschenes, J. (2010). The aesthetics
He has been a visiting scholar at the Marketing Group, Haas
of luxury fashion, body and identify information. Journal of Consumer School of Business, University of California, Berkeley, and has
Psychology, 20(3), 459–470. published in journals such as Journal of International Marketing,
Volonté, P. (2019). The thin ideal and the practice of fashion. Journal of Journal of Business Research, International Marketing Review,
Consumer Culture, 19(2), 252–270.

n, T. (2004). The effect of word of mouth on
European Journal of Marketing, Journal of Consumer Marketing,
Von Wangenheim, F., & Bayo
services switching. European Journal of Marketing, 38(9/10), 1173– Journal of Consumer Behaviour, International Journal of Con-
1185. sumer Studies, Journal of Retailing and Consumer Services, Com-
Voorhees, C. M., Brady, M. K., Calantone, R., & Ramirez, E. (2016). Discrim- puters in Human Behavior, and Psychology & Marketing.
inant validity testing in marketing: An analysis, causes for concern, and
 14791838, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cb.2274 by The University Of Texas Rio Grande Vallley, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
VAZQUEZ ET AL. 19

SUPPORTING INF ORMATION

Lorena Siliceo is an HR executive at SumUp, France and an affili-
Additional supporting information can be found online in the Support-
ated lecturer of management at the Anahuac University, Mexico.
ing Information section at the end of this article.
She holds a Master of Science in International Business from the
University of Edinburgh, UK and has conducted research related
to addressing inequality and diversity as consultant to The City of How to cite this article: Vazquez, E. E., Felix, R., & Siliceo, L.
Edinburgh Council, UK. Mrs. Siliceo has extensive experience in (2023). Female advertising models and their body mass index
the fashion, retailing, and fintech industries serving in executive (BMI): How does BMI similarity affect perceived product
positions for startups and scaleups. Mrs. Siliceo also served at quality? Journal of Consumer Behaviour, 1–19. https://doi.org/
Regnum Christi where she managed multiple private schools in 10.1002/cb.2274
Latin America. Her research interests revolve around people ana-
lytics, entrepreneurial finance, as well as diversity and inclusion.