Communication Monographs

ISSN: 0363-7751 (Print) 1479-5787 (Online) Journal homepage: http://www.tandfonline.com/loi/rcmm20

Distinguishing technologies for social

interaction: The perceived social affordances of
communication channels scale

Jesse Fox & Bree McEwan

To cite this article: Jesse Fox & Bree McEwan (2017): Distinguishing technologies for social
interaction: The perceived social affordances of communication channels scale, Communication
Monographs, DOI: 10.1080/03637751.2017.1332418

To link to this article: http://dx.doi.org/10.1080/03637751.2017.1332418

Published online: 28 Jun 2017.

Submit your article to this journal

Article views: 85

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=rcmm20

Download by: [University of New Mexico] Date: 02 July 2017, At: 08:02
COMMUNICATION MONOGRAPHS, 2017
https://doi.org/10.1080/03637751.2017.1332418

Distinguishing technologies for social interaction: The

perceived social affordances of communication channels scale
Jesse Foxa and Bree McEwanb
a
School of Communication, The Ohio State University, Columbus, OH, USA; bCollege of Communication,
DePaul University, Chicago, IL, USA

ABSTRACT ARTICLE HISTORY

The concept of affordances in communication technology research Received 3 August 2016
has proven to be heuristically provocative, yet perceived Accepted 3 April 2017
affordances are rarely measured. After extracting commonly cited
KEYWORDS
social affordances from the literature, we developed a measure to Affordances; channel
assess participants’ perceptions of these affordances. The scale selection; computer-
was tested across eight communication channels in two studies mediated communication;
(face-to-face; texting; phone; email; posts on social networking social media; texting; face-to-
sites, specifically Facebook; instant messaging; Skype face communication
videoconferencing; and mobile app Snapchat). A factor structure
was developed in Study 1 and confirmed in Study 2. The resultant
Perceived Social Affordances of Communication Channels Scale
includes 41 items measuring 10 communicative affordances:
accessibility, bandwidth, social presence, privacy, network
association, personalization, persistence, editability, conversation
control, and anonymity. Potential methodological and theoretical
applications are discussed.

As a range of mediated channels have become integral to everyday communication, scho-

lars have speculated about whether differences in these technologies are a significant com-
ponent of message processes. Some treat communication technologies as homogeneous,
lumping all forms of computer-mediated communication (CMC) together or considering
diverse channels in a monolithic fashion (see critiques by Sundar & Bellur, 2011; Walther,
Gay, & Hancock, 2005). Other scholars take this assertion a step further, suggesting com-
munication technologies are functionally indistinguishable from other tools (see
McOmber, 1999). These perspectives overlook the fact that channel attributes vary, and
emerging communication technologies have attributes unlike other interpersonal or
mass communication channels. These affordances enable divergent uses and experiences
within these channels, and as such, they have distinct social implications. For example, a
comment on a social networking site profile has a far greater reach than a text, an email, or
even a face-to-face comment. Moreover, individual perceptions may vary as well; users
may think their comment is visible only to their personal network, to any user with a
profile on the site, or perhaps to anyone on the internet.
An affordance-based approach may resolve several deficiencies in the way channels and
channel effects have been handled in previous scholarship. Several scholars have pushed

CONTACT Jesse Fox [email protected]

© 2017 National Communication Association
2 J. FOX AND B. MCEWAN

back against claims and theorizing based on the channel itself (e.g., Walther et al., 2005),
instead advocating for approaches that account for distinct features (e.g., a variable-cen-
tered approach; Nass & Mason, 1990). Another issue is that assumptions in channel the-
ories are often based on the capabilities of channels at that point in time; both channels
and users’ perceptions can evolve (Ellison & Vitak, 2015). For example, the social identity
model of deindividuation effects (SIDE) suggests that disinhibited communication is more
likely in computer-mediated environments than face-to-face because in CMC channels
available at the time, fewer cues were able to be transmitted and participants were often
anonymous (Lea & Spears, 1991). Current CMC channels, such as social networking
sites and videoconferencing, challenge these original conceptualizations. Further, users
may be able to customize features within these channels, such as changing the privacy set-
tings on one’s social media profile. Thus, considering affordances instead of channels
should offer more nuanced – and durable – theorizing that is more flexible across contexts
(Ellison & Vitak, 2015). Whether or not a channel like Snapchat diminishes in popularity
like MySpace or the WELL, or fades to obsolescence like VideoTex, is less relevant than
what we can learn from affordances such as persistence, personalization, and privacy.
Of course, channel features have not been ignored entirely. Affordances have been inte-
grated in channel-focused theories for decades (e.g., social presence theory, Short, Wil-
liams, & Christie, 1976; media richness theory, Daft, Lengel, & Trevino, 1987) and a
considerable number of studies have identified relationships between channel affordances
and outcomes (e.g., Bazarova, 2012; Lee & Oh, 2012; Tanis & Postmes, 2003; Walther,
1994). Unfortunately, many approaches merely assume affordances differ meaningfully
between channels and they are not suitably operationalized in study design.
Another concern is that when affordances are included, they are treated as objective
properties of the channel and overlook the human experience. Numerous studies have
shown that people may have widely varying perceptions of channel affordances, particu-
larly in emerging media environments (e.g., Carlson & Zmud, 1999; DeAndrea, Van Der
Heide, Vendemia, & Vang, in press; Feaster, 2010; Schouten, Valkenburg, & Peter, 2007;
Tao & Bucy, 2007). For example, James may think choosing an avatar and handle not tied
to his real name on Twitter makes him anonymous, whereas Ashley may find this insuffi-
cient given that geolocated posts, connections within the network, and the content of posts
reveal sufficient clues about one’s identity. Indeed, previous research has found that objec-
tive manipulations of anonymity do not necessarily explain outcomes, whereas percep-
tions of anonymity can (Rains, 2007). Put simply, perceived affordances matter
regarding the use and effects of communication channels.
These issues could be addressed if researchers consistently assessed users’ perceptions
of affordances. The field has lacked validated ways to systematically assess perceptions of
affordances, however, limiting scholars’ ability to make comparisons across channels,
across studies, or between users. Thus, the goal of this project was to develop a measure
to assess a broad range of perceived social affordances across various communication
channels.

The significance of affordances

Gibson (1979) defined affordances as the inherent functional attributes of a particular
object emerging in the relationship between actor and object. According to Gibson,
 COMMUNICATION MONOGRAPHS 3

affordances are action possibilities that require both entities, but exist independent of the
actor’s perceptions. This conceptualization is similar to how media scholars have
described “features” or “attributes” of communication channels (e.g., Daft et al., 1987;
Eveland, 2003; Tao & Bucy, 2007).
In contrast to Gibson (1979), Norman (1990) argued for the importance of assessing
perceived affordances based on the user’s experience rather than the inherent properties
of an object. Users imbue objects with their own interpretations that may not correspond
with the intention of an object’s design. For example, although a gleaming silver tureen
may be designed as an aesthetic award, the award recipient may find it a suitable dog dish.
As noted, affordances are fundamental to many mediated communication theories,
although they are not always conceptualized as such. Medium theory, for example,
“focuses on such characteristics of each medium and on how each medium (or each
type of media) is physically, socially, and psychologically distinct from other media”
(Meyrowitz, 2009, p. 518). Similarly, early theorizing about CMC sought to clarify the
meaningful differences and similarities between CMC and face-to-face communication
(Walther & Parks, 2002). Given CMC at the time was generally text based, much of the
focus was on the lack of affordances to convey certain cues and communicate affective
states (e.g., Daft et al., 1987; Short et al., 1976).
As communication technologies become more widely adopted, researchers have delved
more deeply into variation among channels. Technological affordances identified as
having meaningful implications for human interaction have been labeled social affor-
dances (e.g., Bradner, Kellogg, & Erickson, 1999; Wellman et al., 2003), sociotechnical
affordances (e.g., Bazarova, 2012), or communicative affordances (Hutchby & Barnett,
2005; Parks, 2008). Hogan (2009) challenged previous applications of this term by
Wellman et al. and others, noting it typically overlooked individual perceptions. Hogan
redefined social affordances as “the perceptual cues that connote aspects of social structure
to individuals thereby creating a functional difference for the individual” (p. 27). Although
some scholars have argued for a more Gibsonian approach to affordances (e.g., Sundar,
2008), several scholars have argued that it is inadequate for communication research.
Schmitz and Fulk (1991) noted that treating a characteristic-like media richness as an
objective feature of the medium rather than considering users’ perceptions is a “specifica-
tion error” that “reduces our ability to accurately predict communication behavior in
organizations” (p. 491). Similarly, Feaster (2010) argued this approach “can in effect
exclude the communicator from the communication situation” (p. 119). Given our scale
is intended to be used in communicative contexts, we adhere to Hogan’s (2009) conceptual
definition and focus on perceptions.

Existing approaches to measuring affordances

Affordances are not entirely overlooked in studies, but there are several shortcomings in
their operationalization. Many researchers treat affordances as fixed, invariant capacities
of a medium. For example, experimentalists often operationalize an affordance as “low”
and “high” or “absent” and “present” rather than looking at degrees or accounting for
users’ interpretations (see critiques by Feaster, 2010; Fulk, Schmitz, & Steinfield, 1990;
Schouten et al., 2007; Walther, 2009). Norman (1990) and Gaver (1991) argued that ignor-
ing users’ perceptions is problematic because it leads to the creation of less intuitive and
4 J. FOX AND B. MCEWAN

sound design. From the perspective of communication research, this approach is also
problematic.
Given the persistence of digital divides and differences in media use (Pearce & Rice,
2013; Pew Research Center, 2015), it is presumptive to believe users’ perceptions are
aligned with the researcher’s. Individual differences such as cognitive capacity, media lit-
eracy, or physical limitations may influence a user’s ability to evaluate affordances of a
channel the way the researcher expects. People can perceive affordances that are not
inherent to an object or fail to perceive affordances that exist. For example, online
forum users may feel their posts are anonymous, but many sites record IP addresses,
which makes users identifiable. The mismatch between user perception and the object’s
inherent properties yields false affordances (Gaver, 1991). Another concern is hidden affor-
dances, in which users do not perceive an inherent affordance (Gaver, 1991). Someone
might resist using Twitter because of its public nature, not realizing privacy settings are
available. In both of these cases, it is likely users’ behaviors are better predicted by their
perceptions of affordances rather than these sites’ actual affordances.
Collectively, these arguments suggest that conceptualizing affordances as objectively
present or not, high or low, may overlook some crucial and theoretically important vari-
ation in the ways people experience communication channels. Such an approach may lead
to attributions toward a channel rather than clarifying the mechanisms driving effects.
Following many communication technology scholars (e.g., Feaster, 2010; Fulk et al.,
1990; Schmitz & Fulk, 1991; Schouten et al., 2007; Steinfield, 1986; Tao & Bucy, 2007),
we agree that assessing individuals’ perceptions of these affordances will provide greater
insight into social interaction.

The need for a unified measure of perceived affordances

At this time, a validated method of assessing perceived social affordances is necessary to
facilitate more consistent, standardized tests across channels. Often technologies are the-
orized to be inherently higher on a specific affordance, but direct comparisons between
channels are rarely made (Walther, 2013). As Walther (2013) argued, these assumptions
may lead to “type I technology errors” in which researchers attribute findings to the fea-
tures or affordances of a technology without conducting suitable comparisons to other
contexts. Given many theoretical claims are based on assumed differences between chan-
nels, it is crucial to develop measures that test these assumptions and ensure that con-
clusions are valid. Findings that have been attributed to a channel may be the effects of
a perceived affordance independent of a specific channel.
One issue in making such comparisons is that existing measures are rarely worded flex-
ibly enough to describe multiple channels. Thus, researchers often must adapt or concoct
various measures for each study. Not only are these rarely validated, but this practice has
made it difficult to compare findings across studies systematically as observed differences
may be attributed to measurement variation. As Walther (2013) argued, for affordances to
be theoretically useful, they must “provide a stable basis for comparisons” (p. 191).
In addition to examining multiple channels, a unified scale presents the opportunity to
examine multiple affordances. Methodologically, a scale provides researchers the oppor-
tunity to investigate the co-occurrence of, and potential interactions between, perceived
affordances, which are important to consider both in the natural use of technologies
 COMMUNICATION MONOGRAPHS 5

and in terms of theoretical validity (Eveland, 2003; Walther, 2009). If researchers patch
together various existing measures, it is likely they may not represent clean factors
given that many affordances are closely related.
A final issue is that many “measures” of affordances are not derived using appropriate
scale development techniques. Although items and scale reliability are typically reported,
these are not sufficient indicators of measurement validity. Rarely do these measures
incorporate procedures such as developing item pools; identifying and refining factor
structure; establishing multiple aspects of validity; or testing items across multiple
samples (DeVellis, 2017). The assessment of perceived affordances requires a unified, psy-
chometrically sound measure.

Social affordances of communication channels

To identify the social affordances examined by communication researchers, we conducted
a thorough literature review on channel, media, and technological features, attributes, and
affordances and also delved into channel-related theories. We focused on (1) affordances
relevant to human interaction (as opposed to human–computer interaction, interacting
alone with a device or interface) that (2) applied across multiple channels including
face-to-face given it is frequently theorized as a basis for comparison. Moreover, we
selected affordances that (3) have been posited as mechanisms driving channel selection,
use, and effects.
One of the most frequently cited affordances is bandwidth, the breadth of social cues
potentially transmitted in a channel (Daft et al., 1987; Rice & Steinfield, 1994; Short
et al., 1976; Wellman et al., 2003). Nonverbal cues in particular can be filtered out or
over-emphasized in mediated channels. These cues are enabled by the structural features
of a channel such as text, graphical icons, photographs, audio, video, or other sensory
information such as touch and smell (similar to conceptualizations of modality or multi-
mediality, e.g., Culnan & Markus, 1987; Gaver, 1991; Reid & Reid, 2010; Schrock, 2015;
Sundar, 2008). Bandwidth is fundamental to theories such as SIDE and social information
processing theory (SIP). A lack of cues may lead to misunderstandings or negative inter-
actions (SIDE; Lea & Spears, 1991). Consequently, users often try to replace or simulate
them (SIP; Walther, 1996). For example, text can be enriched through the use of graphical
icons such as emojis or accompanying photographs.
Similarly, channels vary in their ability to promote social presence, or the feeling that
interactants are near and sharing the same experience together (Bradner et al., 1999;
Clark & Brennan, 1991; Daft et al., 1987; Rice & Steinfield, 1994; Short et al., 1976; Wil-
liams & Rice, 1983). Both social presence theory (Short et al., 1976) and media richness
theory (Daft et al., 1987) highlight the role of this affordance. Many computer-mediated
channels are asynchronous, however, which can minimize feelings of social presence.
Rather than providing instant feedback, there is a lag between message transmission,
receipt, and subsequent response (Bradner et al., 1999; Culnan & Markus, 1987; Rice &
Steinfield, 1994). This delay, however, can be beneficial as it may enable more strategic
monitoring of one’s interactions; this is a fundamental claim of approaches such as the
hyperpersonal model (Walther, 1996).
As such, asynchronicity is associated with a form of expressive or communicative
control (Feaster, 2010; O’Sullivan, 2000; Reid & Reid, 2010; Schouten et al., 2007; also
6 J. FOX AND B. MCEWAN

referred to as interaction management, Walther & Parks, 2002). In contrast to face-to-face,

users of mediated channels often have greater control over the rate and nature of their
response. Conversational control is the ability to manage the mechanics of an interaction,
such as ending a conversation or regulating turn-taking (Feaster, 2010). Different channels
also afford users different levels of informational control regarding what content they wish
to express or withhold within interactions (Feaster, 2010). This macro-level of control is
also related to editability, or the capacity to revise specific messages before sharing (Rice,
1987; Treem & Leonardi, 2013; Walther, 1996).
Persistence (akin to archiving, Resnick, 2001; durability, Parks, 2008; recordability,
Toma, Hancock, & Ellison, 2008; and reviewability, Clark & Brennan, 1991) describes
the relative permanence or ephemerality of communication (boyd, 2011; Culnan &
Markus, 1987; Treem & Leonardi, 2013). Because digital material is easily captured,
saved, replicated, and recirculated, it may be accessible long after the initial interaction
and difficult to remove permanently. Persistence has been suggested as a possible expla-
nation for why the spiral of silence theory may operate differently online, as users
might consider that their political expressions may remain observable and be held
against them in the future (Stoycheff, 2016).
The affordance of network association (also referred to as connectivity or linkage)
enables group members, no matter how disparate or geographically distant, to identify
other members. Often, users can view each other’s content or connect to another
member through a common node or “friend” (boyd & Ellison, 2007; Fox & Moreland,
2015; Parks, 2008; Treem & Leonardi, 2013). These links form a network of known and
latent ties (Haythornthwaite, 2002). Markers such as lists or links make it easy to identify
or contact other parties in another’s social network without them having to be present
during an interaction. Concepts such as context collapse, when social circles converge
that are normally maintained separately, are characterized by the affordance of association
(Ellison & Vitak, 2015). Privacy, or alternately visibility, is the degree to which communi-
cation is apparent to many others, whether intended or not (boyd & Ellison, 2007; Culnan
& Markus, 1987; Steinfield, 1986; Treem & Leonardi, 2013; Williams & Rice, 1983). Chan-
nels like texting, phone, or email are perceived as relatively private as information is shared
with specific recipients. In contrast, information posted on a public website is visible to
anyone with the link or search skills to find it.
The intended audience within a channel may be associated with personalization or tai-
loring, defined as the ability to direct a message to a specific individual or group (Daft
et al., 1987; Wellman et al., 2003). According to media richness theory, richer channels
enable personalization and, in turn, facilitate more efficient communication (Daft et al.,
1987). Personalization may be accomplished easily in one-to-one interactions (whether
face-to-face or mediated) but be more difficult in one-to-many (i.e., mass or massperso-
nal) communication. In the same way that the receiver can be specified, channels vary in
whether the sender is recognized. Anonymity, also referred to as identifiability, is the
degree to which users feel their real names or true identities can be concealed in a
channel regardless of how public or private their communication may be (Lea &
Spears, 1991; Resnick, 2001; Rice & Steinfield, 1994). Theoretical approaches such as
SIDE suggest that anonymity drives feelings of depersonalization and disinhibited behav-
ior (Lea & Spears, 1991).
 COMMUNICATION MONOGRAPHS 7

Finally, the accessibility of a channel is the capability of easily achieving or reaching

communication, regardless of time, place, structural limitations, technological literacy,
or other constraints (Bradner et al., 1999; Culnan & Markus, 1987; Fox & Moreland,
2015; Reid & Reid, 2010; Wellman et al., 2003; similar to convenience, Walther, 1996,
and ease of use, Ledbetter, 2009; Parks, 2008). Accessibility is fundamental to phenomena
such as the diffusion of innovations (Rogers, 1962) or the digital divide (Pearce & Rice,
2013). Mobile devices, and particularly smartphones, have increased the availability of
many channels, especially with the expanding technological infrastructure (Reid & Reid,
2010; Schrock, 2015). As such, users may transcend the temporal and geographical con-
straints and reach out to others at any time and from any place using a connected
digital device.

Current aims and item development

Although existing scales measure experiences with technology, currently there is no scale
designed to assess perceptions of multiple social affordances associated with multiple com-
munication channels. Although Sundar and Limperos (2013) situated their measure
around affordances, it focuses on motivations of users independent of a communicative
context, such as the novelty of a technology or ease of navigation within an interface.
Other measures we identified were specific to a channel or type of channel (e.g., Ledbetter’s
(2009) Measurement of Online Communication Attitudes [MOCA]), and some items were
not applicable or flexible to other channels, particularly face-to-face communication. Some
identified measures featured items that included multiple affordances (e.g., Kelly and
Keaten’s (2007) Affect for Communication Channels Scale [ACCS]). Because we wanted
to make a modular scale in which different affordances could be assessed individually or
together, it was important to ensure that our items were distinct and would not cross load.
With these guidelines in mind, we developed 66 items to assess 12 social affordances
identified within the literature. We selected affordances that were fundamental to
channel-based theorizing, mentioned across multiple sources, and offered heuristic
value: accessibility (4 items), bandwidth (6 items), social presence (5 items), privacy (5
items), network association (6 items), personalization (5 items), persistence (6 items), edit-
ability (4 items), information control (5 items), conversation control (4 items), anonymity
(7 items), and synchronicity (9 items). See Table 1 for all original items.
For comparison, we assessed individuals’ perceptions of these 12 affordances across 5
channels commonly examined in channel-based communication research (e.g., Kelly &
Keaten, 2007; Ledbetter, 2014; Ramirez, Dimmick, Feaster, & Lin, 2008; Ruppel &
Burke, 2015): face-to-face, phone, texting, email, and social networking sites (SNSs).
Because many SNSs feature several routes for communication, we specifically focused
on Facebook timeline posts (i.e., content posted on one’s own page).

Study 1
Sample and procedure
After IRB approval was obtained, Mechanical Turk workers (N = 875) who identified as
English speaking U.S. residents were recruited to participate in exchange for $2.
8 J. FOX AND B. MCEWAN

Table 1. Final measurement model with scale items.

β Β
Accessibility (α = .79, .79) CFA: χ 2/df = 1.35, CFI = .996, SRMR = .02, RMSEA = .05 (.00, .17)
This channel is convenient. .83 .80
It is easy for me to access this channel. .78 .78
This channel makes it easy to get a message to someone. .64 .65
This channel makes it easy to communicate.
Bandwidth (α = .90, .89) CFA: χ 2/df = 1.56, CFI = .988, SRMR = .03, RMSEA = .06 (.00,.11)
This channel allows me to convey emotion. .93 .86
This channel allows me to express emotion. .91 .80
This channel allows me to receive cues about how the other person is feeling. .81 .84
In this channel, I can say not just what I want to say, but how I want to say it. .72 .76
This channel allows me to express things nonverbally.
This channel allows me to receive cues about how the other person is feeling.
Social presence (α = .91, .89) CFA: χ 2/df = 1.74, CFI = .99, SRMR = .03, RMSEA = .06 (.00, .14)
This channel makes it seem like the other person is present. .91 .80
This channel makes it feel like the person I’m communicating with is close by. .90 .86
This channel makes it feel like other people are really with me when we communicate. .85 .84
This channel allows me to determine if someone is really “there” when communicating. .80 .76
Using this channel makes me feel fairly distant from the other person.a,b .64
Privacy α = (.84, .85) CFA: χ 2/df = 1.18, CFI = .997, SRMR = .03, RMSEA = .03 (.00, .12)
This channel helps keep my communication private. .91 .87
My communication is private via this channel. .91 .85
I expect my communication to be private via this channel. .89 .71
It is difficult to know exactly who can observe communication taking place in this channel.a
This channel makes it easy for others to share in communication that I intend to keep private.a
Network Association α = (.91, .83) CFA: χ 2/df = 1.73, CFI = .980, SRMR = .04, RMSEA = .07 (.00, .12)
In this channel, members of our social networks can easily join or interrupt our interaction. .89 .72
Communication with someone through this channel makes our connection apparent to other network .87 .77
members.
Communicating through this channel allows many members of our social network to be part of our interaction. .80 .66
This channel makes my relationships to other people visible. .80 .63
This channel makes it easy for others to identify other people I am connected to. .76 .75
This channel allows me to connect with multiple other communicators.
Personalization (α = .80, .78) CFA: χ 2/df = 1.38, CFI = .98, SRMR = .03, RMSEA = .05 (.00, .12)
This channel allows me to focus my message on a specific person. .83 .81
This channel allows me to adapt my message for a specific receiver. .75 .70
This channel allows me to address my communication only to certain people. .73 .59
This channel allows me to personalize my message. .62 .69
In this channel, I cannot target messages to only a specific person.a,b .51
Persistence (α = .95, .89) CFA: χ 2/df = 1.86, CFI = .98, SRMR = .04, RMSEA = .07 (.00, .12)
This channel keeps a record of communication that I can go back and look at. .96 .91
I can retrieve past messages in this channel. .94 .87
This channel keeps a record of communication that can last long after the initial communication. .93 .85
Communication in this channel exists long after the initial interaction is finished. .82 .65
Communication in this channel exists only in the moment.a,b .84
This channel keeps no permanent record of what I say or do.a,b .77
Editability (α = .92, .87) CFA: χ 2/df = .37, CFI = 1.00, SRMR = .01, RMSEA = .00 (.00,.11)
This channel allows me to create a message and delete it before I communicate it. .92 .86
This channel allows me to edit the message I want to communicate before I actually communicate it. .87 .89
This channel allows me to carefully craft my message before sending it. .87 .82
If I make a mistake when creating a message in this channel I can change it before my receiver gets it. .70 .62
Conversation Control (α = .77, .78) CFA: χ 2/df = 2.39, CFI = .99, SRMR = .03, RMSEA = .09 (.00, .20)
This channel allows me to control the duration of the conversation. .75 .76
I can control the amount of time I invest in a conversation through this channel. .66 .71
This channel allows me to end an interaction if I need to do so. .66 .70
This channel allows me to regulate the flow of communication with others. .64 .62
Anonymity (α = .93, .95) CFA: χ 2/df = 1.07, CFI = .999, SRMR = .02, RMSEA = .02 (.00, .08)
This channel can make me anonymous to the person I am communicating with. .84 .89
The channel allows people to remain anonymous or unidentifiable if they want to. .84 .87
When using this channel, I can take on another identity if I want to. .84 .84
This channel can mask my true identity when communicating. .81 .88
When I communicate through this channel, the receiver doesn’t necessarily know it’s me. .80 .85

(Continued )
 COMMUNICATION MONOGRAPHS 9

Table 1. Continued.
β Β
You can’t necessarily tell who is communicating through this channel. .78 .84
This channel makes it difficult to conceal one’s identity when communicating.a,b .70
Information Control (α = .61) CFA: χ 2/df = .75, CFI = 1.00, SRMR = .02, RMSEA = .00 (.00, .14)
This channel allows me to control how much information I am disclosing both verbally and nonverbally.
This channel allows me to control how much information the other person receives.
This channel allows me to avoid topics that I don’t want to discuss.
This channel allows me to control how much information I receive from others.
This channel allows me to filter out information if I want to.
Synchronicity (α = .85) CFA: χ 2/df = 1.67, CFI = .95, SRMR = .05, RMSEA = .06 (.03, .09)
The messages I send can be observed in real time – that is, the receiver can observe me while I am communicating.
There is a lapse of time in between the messages I send and the other person receiving the message.
This channel allows for instant communication.
It can be a long time before you get a response using this channel.a
This channel allows communicators to quickly send messages back and forth.
I should not expect to get my message right away when using this channel.a
When using this channel, I expect the other person will respond quickly.
This channel allows me to give and receive timely feedback.
This channel allows for back and forth interaction. a
Scale items: a indicates reverse scoring. Items in italics are not included in the final scale, including the entire information
control and synchronicity factors. b indicates items retained in Study 1 but removed from the final scale in Study 2.
Alphas: The reported Study 1 alpha includes all Study 1 items. Study 2 alphas are in boldface. CFAs: Reported CFAs are the
unidimensional CFA.

Consenting participants were directed to an online survey. After completing demo-

graphics measures, participants were randomly assigned to one of five channels. They
responded to the affordance and validation items on a fully-labeled 7-point scale (1 =
strongly disagree; 7 = strongly agree).
Given we did not prescreen for use of all five media, some attrition occurred due to
channel assignment. Participants who claimed they did not use the particular channel
they were assigned were excluded (n = 30). We also asked participants to indicate their
channel assignment and removed those who indicated the wrong channel (n = 37). An
additional 7.4% missed at least one of the eight attention check questions within the
survey. Individuals who missed two or more attention checks (n = 4) were removed from
the data set. The remaining data from individuals who missed only one attention check
(n = 54) were examined more closely and deemed acceptable; t-tests between those who
missed one and those who missed no attention checks revealed no significant differences
on any demographic or affordance variable (p > .05). After the data cleaning process, par-
ticipants were still relatively evenly distributed across channels: face-to-face (n = 159), tele-
phone calls (n = 169), texting (n = 166), Facebook posts (n = 151), and email (n = 159).
The final sample (N = 804) consisted of 401 male and 403 female participants aged 19 to
69 (M = 35.21; SD = 10.60). Participants indicated their race/s or ethnicity/ies as: 81%
White/Caucasian, 8% Black/African-American, 6.2% Asian/Asian-American, 5.7%
Latino/a, and 1.5% identified as other groups. In terms of education, 47% reported
earning a college degree, 28.9% had completed some college, 13.2% had completed high
school, 10.2% had earned a graduate degree, and 0.6% had not completed high school.

Unidimensional confirmatory factory analyses (CFA)

A first-order CFA was performed for each affordance to determine its unidimensionality.
Model fit was considered adequate if the χ 2/df was less than 3, the CFI was greater than .95,
10 J. FOX AND B. MCEWAN

the SRMR was below .08 (Hu & Bentler, 1999), and the RMSEA was below .08 (taking into
account the associated confidence interval, Browne & Cudeck, 1993; Kline, 2016). Items
were examined if they did not fit the unidimensional model. Upon further consideration,
some of these items were determined to be vague or worded in a way that was awkward for
a particular channel and were removed. Initial models for synchronicity and anonymity
were acceptable. Three items each were trimmed from accessibility and social presence
and two from bandwidth and conversational control. One item each was trimmed from
privacy, network association, personalization, editability, and information control. The
complete list of items, unidimensional factors after item trimming, and final model fit stat-
istics can be found in Table 1.

CFA for the full measurement model

A confirmatory factor model was estimated with all 12 factors using only the items
retained from the unidimensional models. We anticipated a lower CFI due to the complex-
ity of the model and set the model fit criteria for CFI at .90 (see Kline, 2016; Marsh, Hau, &
Wen, 2004, for a discussion of setting a priori model fit criteria). We retained the same
model fit criteria as the unidimensional models for the other three fit criteria (χ 2/df < 3.0,
SRMR < .08, RMSEA < .08). The initial model fit was inadequate on three of four selected
fit indices, χ 2/df = 3.56, CFI = .87, SRMR = .09, RMSEA = .056 (.055, .058) and was evaluated
for improvement.
The latent variable estimate for the synchronicity factor revealed serious instability
within the factor, as all items cross-loaded with multiple other factors. Thus, the entire
factor was removed from the model. This improved model fit, although it was still some-
what inadequate on one fit index, χ 2/df = 3.28, CFI = .90, SRMR = .08, RMSEA = .053
(.052, .055). Individual items were evaluated for both statistical cross-loading (items
were dropped based on low loadings on their own factor while cross-referencing loadings
onto other factors) as well as semantic content and removed based on this assessment. One
item was removed from each of the factors accessibility and network association. Two
items were removed from bandwidth and privacy.
The final model achieved an acceptable level of fit, χ 2/df = 2.84, CFI = .93, SRMR = .07,
RMSEA = .048 (.046, .050). The individual factors were then assessed for inter-item
reliability. All factors met the standard cutoff for Cronbach’s alpha (α = .70) with the
exception of information control (α = .61). An investigation of the items by specific chan-
nels suggested that overall the items were unreliable because the types of information
control (avoiding, filtering, receiving) may be different for different channels. The infor-
mation control items were removed from the overall model, and the model fit remained
essentially the same, χ 2/df = 2.83, CFI = .94, SRMR = .06, RMSEA = .048 (.046, .050).
Descriptive statistics for each factor are included in Table 1. See Table 2 for correlations
among factors.

Validation
We selected one existing measure for each our factors to examine convergent validity with
variables similar to each factor. We validated accessibility with the ease of use subscale
of the MOCA, given both scales concern the availability and convenience of channels
 COMMUNICATION MONOGRAPHS 11

Table 2. Correlations between PSACCS factors and validation scales.

1 2 3 4 5 6 7 8 9 10
1. Accessibility –
2. Bandwidth .06 –
3. Privacy .22* .27* –
4. Connectivity −.02 .21* −.34* –
5. Persistence .31* −.49* −.11* .15* –
6. Editability .33* −.40* −.05 .12* .73* –
7. Conv. Control .42* .16* .30* −.07 .11* .21* –
8. Personalization .39* .30* .45* −.25* −.06 −.04 .48* –
9. Social Presence −.01 .74* .31* .25* −.50* −.43* .11* .25* –
10. Anonymity .13* −.31* .04 .08 .46* .46* .11* −.07 −.29* –
11. Ease of Use .65* −.35* .06 −.07 .67* .61* .38* .18* −.40* .38*
12. Clarity .26* .43* .39* −.20* −.26* −.21* .34* .41* .36* −.22*
13. Privacy Concern −.02 −.36* −.39* .33* .46* .42* −.17* −.33* −.37* .28*
14. Connectivity* −.08 .24* −.33* .63* −.04 .06 −.11* −.22* .27* −.01
15. Deep Profiling .09 −.30* −.41* .56* .59* .58* −.11* −.34* −.31* −.22*
16. Affect for Channel .44* −.42* −.02 .02 .74* .74* .31* .12* −.48* .38*
17. Info Control .48* .14* .24* −.14* .16* .22* .67* .41* .04 .14*
18. Int. Credibility .15* .44* .49* −.23* −.37* −.31* .28* .40* .42* −.27*
19. Attentiveness .22* .58* .32* −.16* −.37* −.33* .39* .59* .49* −.30*
20. Presence .00 .70* .28* −.03 −.59* −.50* .16* .27* .72* −.40*
21. Online Anonymity .03 .03 .09 .12* .10* .19* .03 −.18* −.03 .34*
11 12 13 14 15 16 17 18 19 20
1. Accessibility
2. Bandwidth
3. Privacy
4. Connectivity
5. Persistence
6. Editability
7. Conv. Control
8. Personalization
9. Social Presence
10. Anonymity
11. Ease of Use –
12. Clarity .09 –
13. Privacy Concern .20* −.42* –
14. Connectivity* −.04 −.03 .24* –
15. Deep Profiling .33* −.24* .50* .43* –
16. Affect for Channel .74* .01 .28* .01 .45* –
17. Info Control .50* .47* −.15* .02 .03 .49* –
18. Int. Credibility .02 .66* −.43* −.07 −.33* −.09 .36* –
19. Attentiveness −.02 .55* −.40* .04 −.32* −.09 .43* .48* –
20. Presence −.21* .46* −.38* .20* −.31* −.36* .20* .53* .58* –
21. Online Anonymity .16* −.02 .09 .16* .20* .18* .20* .03 −.13* −.01
Notes: Bolded items are correlations with PSACCS factors and their associated validation scales. Validation scale correlations
were conducted with Study 1 data and factors. Correlations between factors were conducted with data from both studies.
*p < .001.

(α = .69, M = 5.24, SD = 1.24; Ledbetter, 2009); bandwidth with the clarity factor of Spitz-
berg’s (2006) CMC competence scale (α = .78, M = 5.11, SD = 1.04), given that the range of
available cues is associated with message ambiguity (e.g., SIP, Walther, 1996); privacy with
Buchanan, Paine, Joinson, and Reips’s (2007) measure of online privacy concern (α = .93,
M = 3.18, SD = 1.53); network association with Boster, Kotowski, Andrews, and Serota’s
(2011) connectivity scale (α = .85, M = 4.03, SD = 1.45); persistence with Suh and
Wagner’s (2013) deep-profiling items (α = .64, M = 3.75, SD = 1.52); editability with the
increased preparation and control subscale of Kelly and Keaten’s (2007) ACCS (α = .93,
M = 4.94, SD = 1.27); conversation control with Feaster’s (2010) information control
12 J. FOX AND B. MCEWAN

scale (α = .76, M = 4.94, SD = 1.43); personalization with the attentiveness factor of Spitz-
berg’s (2006) CMC competence scale (α = .73, M = 5.79, SD = .86); social presence with the
social richness subscale of Lombard, Ditton, and Weinstein’s (2009) Temple Presence
Inventory (α = .91, M = 5.38, SD = 1.14); and anonymity with Ritter’s (2014) perception
of online anonymity subscale (α = .59, M = 3.30, SD = .94), taken from her measure of
cybersexual harassment.
As seen in Table 2, the affordance factors were strongly and positively correlated with
their validation measures with two exceptions. The affordance of privacy was negatively
correlated with the privacy concern validation measure, but this was anticipated as the
more private a user perceives the channel, the less concerned they should be about
privacy issues while using that channel. The second exception was a relatively low cor-
relation between the anonymity affordance and anonymity validation measure, r(804)
= .34, p < .001. The online anonymity validation measure had low reliability, however;
the instability of the validation measure may have reduced the strength of the
correlation.
Comparing our measures across channels, the means generally reflect differences that
have been observed or theorized in previous scholarship. ANOVAs revealed that face-to-
face had significantly greater bandwidth and social presence, and significantly lower anon-
ymity, than all other channels (all ps < .01). The phone had the second-highest bandwidth
and social presence, and was significantly higher (p < .01) than all of the computer-
mediated channels (i.e., texting, email, and Facebook posts; all ps < .01). Facebook timeline
posts, a masspersonal channel, were significantly lower than all other channels on privacy
and personalization but also significantly higher on network association (all ps < .01).
These results from Study 1 can be viewed in Supplement Table 5 (http://www.
breemcewan.com/psaccs.html).

Study 2
After the factor structure was established and tested across channels in Study 1, we col-
lected a second sample to confirm the factor structure across media that existing research
has attributed with different affordances. We included the 10 affordances that emerged in
Study 1 (accessibility, bandwidth, social presence, privacy, network association, persona-
lization, persistence, editability, conversation control, and anonymity).
Additionally, we wanted to probe some affordances that appeared closely related in
Study 1 to ensure that they are indeed perceived as conceptually distinct by users. For
example, editability and persistence were rather highly correlated for our selected chan-
nels; both were low for face-to-face and phone, and higher for the CMC channels of
texting, Facebook posts, and email. Thus, we decided to test the measure with Snapchat,
which has similar functionality to texting, but is noted for deleting messages after they
have been received. Moreover, given that face-to-face is often heralded as an unparalleled
channel for many affordances such as bandwidth and social presence, we wanted to
examine participants’ perceptions of the closest approximation currently available: video-
conferencing. In addition to Snapchat and video conferencing, we examined instant mes-
saging given its inclusion in previous research (e.g., Ledbetter, 2009; Ramirez et al., 2008)
and also retained face-to-face communication.
 COMMUNICATION MONOGRAPHS 13

Sample and procedure

Adult participants (N = 680) were recruited through a Qualtrics panel, which advertised the
study opportunity to eligible members. The final sample included 485 women, 192 men, and
three individuals who did not report gender aged 18 to 79 (M = 39.11, SD = 15.54). Partici-
pants indicated their race/s or ethnicity/ies as: 78.24% White/Caucasian, 10% Black/
African-American, 6.12% Asian/Asian-American, 6.91% Latino/a, and 1.32% identified as
other groups. In terms of education, 40.9% reported earning a college degree, 27.8% had com-
pleted some college, 17.1% had completed high school, 12.5% had earned a graduate degree,
and 1.8% had not completed high school. Participants were asked if they used each channel
and then were randomly assigned to a channel they reported that they had used (face-to-face,
n = 198; Skype videoconferencing, n = 182; Snapchat, n = 171; instant messaging, n = 129).

Confirmatory factor analysis for Study 2

A confirmatory factor analysis was conducted to determine if the model fit from Study 1 was
appropriate for the data in Study 2. A confirmatory factor model was estimated with all 10
factors that emerged in Study 1. The initial model fit was marginally inadequate, χ 2/df =
3.09, CFI = .90, SRMR = .09, RMSEA = .055 (.053, .058). One item was removed from per-
sonalization and one item from social presence which rendered an adequately fitting model,
χ 2/df = 2.54, CFI = .93, SRMR = .07, RMSEA = .048 (.045, .050). However, three items (two
from persistence and one from anonymity) had low factor loadings (under .50) and were
subsequently removed. The final model fit was χ 2/df = 2.48, CFI = .94, SRMR = .06,
RMSEA = .047 (.044, .049). Cronbach’s alphas for all factors are included in Table 1.
The new model fit was then applied to the Study 1 data, resulting in marginal improve-
ment to the model fit, χ 2/df = 2.81, CFI = .94, SRMR = .06, RMSEA = .047 (.045, .050). This
final set of items composes the Perceived Social Affordances of Communication Channels
Scale (PSACCS) and is visible in Table 3 along with instructions and scale points. Sup-
plement Table 5 provides correlations for each factor using the final set of items and cal-
culated by channel for both Study 1 and Study 2 data.

Discriminant validity
To assess discriminant validity, we conducted channel comparisons using the final version
of the scale. As seen in Table 4, ANOVAs compared each affordance across all eight chan-
nels using both Study 1 and Study 2 data. Supplement Table 6 also provides comparisons
using solely Study 2 data (http://www.breemcewan.com/psaccs.html). Each ANOVA was
significant, and post hoc Scheffe tests indicated significant differences between at least
two channels on each affordance suggesting that the PSACCS was able to discriminate
between different media. Moreover, the new channels performed in largely predictable
ways. Snapchat was seen as significantly less persistent than other forms of CMC such as
texting, email, IM, and Facebook. Videoconferencing was perceived as providing higher
bandwidth and social presence than all other CMC channels, but not as high as face-to-
face communication. Instant messaging was considered the most anonymous channel, fol-
lowed by email. Given that these are predominantly text-based forms of interaction, users
are not physically visible, and typically users may choose any handle or avatar they wish,
these channels are perceived as offering fewer identifying cues than other channels.
14 J. FOX AND B. MCEWAN

Table 3. The Perceived Social Affordances of Communication Channels Scale.

Accessibility (α = .79, .79)
This channel is convenient.
It is easy for me to access this channel.
This channel makes it easy to get a message to someone.
Bandwidth (α = .90, .89)
This channel allows me to convey emotion.
This channel allows me to express emotion.
This channel allows me to receive cues about how the other person is feeling.
In this channel, I can say not just what I want to say, but how I want to say it.
Social Presence (α = .83, .89)
This channel makes it feel like the other person is present.
This channel makes it feel like the person I’m communicating with is close by.
This channel makes it feel like other people are really with me when we communicate.
This channel allows me to determine if someone is really “there” when communicating.
Privacy α = (.93, .85)
This channel helps keep my communication private.
My communication is private via this channel.
I expect my communication to be private via this channel.
Network Association α = (.83, .83)
In this channel, members of our social networks can easily join or interrupt our interaction.
Communication with someone through this channel makes our connection apparent to other network members.
Communicating through this channel allows many members of our social network to be part of our interaction.
This channel makes my relationships to other people visible.
This channel makes it easy for others to identify other people I am connected to.
Personalization (α = .78, .78)
This channel allows me to focus my message on a specific person.
This channel allows me to address my communication only to certain people.
This channel allows me to personalize my message.
Persistence (α = .84, .89)
This channel keeps a record of communication that I can go back and look at.
I can retrieve past messages in this channel.
This channel keeps a record of communication that can last long after the initial communication.
Communication in this channel exists long after the initial interaction is finished.
Editability (α = .87, .87)
This channel allows me to create a message and delete it before I communicate it.
This channel allows me to edit the message I want to communicate before I actually communicate it.
This channel allows me to carefully craft my message before sending it.
If I make a mistake when creating a message in this channel I can change it before my receiver gets it.
Conversation Control (α = .79, .78)
This channel allows me to control the duration of the conversation.
I can control the amount of time I invest in a conversation through this channel.
This channel allows me to end an interaction if I need to do so.
This channel allows me to regulate the flow of communication with others.
Anonymity (α = .92, .95)
This channel can make me anonymous to the person I am communicating with.
The channel allows people to remain anonymous or unidentifiable if they want to.
When using this channel, I can take on another identity if I want to.
This channel can mask my true identity when communicating.
When I communicate through this channel, the receiver doesn’t necessarily know it’s me.
You can’t necessarily tell who is communicating through this channel.
Notes: Participants should respond on a fully-labeled 7-point scale (1 = Strongly disagree; 2 = Disagree; 3 = Slightly
disagree; 4 = Neither agree nor disagree; 5 = Slightly agree; 6 = Agree; 7 = Strongly agree). When using the scale,
researchers may wish to replace “this channel” with the name of the particular channel to improve readability.
Study 2 alphas are in boldface.

Discussion
For decades, communication researchers have used affordances and related concepts to
predict aspects of channel use, yet there is no standard, validated measure for assessing
perceived affordances. The PSACCS comprises 41 items assessing perceptions of 10 dis-
tinct social affordances. The PSACCS can be implemented to assess perceived social
Table 4. ANOVA comparisons of perceived social affordances across channels.
Affordance (F ) F-t-F M (SD) Texting M (SD) Email M (SD) Facebook M (SD) Phone M (SD) Video conferencing M (SD) Snapchat M (SD) IM M (SD)
a e d,e a,b,c a,b,c a,b b,c,d
Accessibility (19.26) 5.56 (1.12) 6.39 (0.81) 6.31 (0.86) 5.72 (1.01) 5.89 (0.87) 5.57 (1.06) 5.92 (0.95) 6.07 (0.91)c,d,e
Bandwidth (175.70) 6.37 (0.71)a 4.10 (1.35)b 4.10 (1.23)b 4.41 (1.22)b,e 5.92 (0.82)c 5.96 (0.80)c 5.58 (1.07)d 4.52 (1.31)e
Privacy (120.24) 5.01 (1.37)a 5.04 (1.35)a 5.29 (1.21)a,c 2.78 (1.51)b 5.65 (1.00)c 5.34 (1.07)a,c 5.03 (1.37)a 4.99 (1.30)a
Net. Assoc. (208.40) 4.47 (1.50)a 2.87 (1.37)b 2.65 (1.19)b 5.51 (0.95)c 2.35 (1.07)b 4.61 (1.14)a 4.17 (1.44)a 4.23 (1.14)a
Persistence (322.65) 2.65 (1.49)a 5.98 (0.93)d,e 6.40 (0.69)e 5.94 (0.83)d,e 2.24 (1.18)a 4.15 (1.38)c 3.33 (1.65)b 5.45 (1.18)d
Editability (238.54) 3.00 (1.59)a 5.73 (0.95)e,f 6.01 (0.85)f 5.48 (0.97)d,e 2.33 (1.28)b 4.25 (1.56)c 5.17 (1.07)d 5.46 (0.86)d,e
Conv. Control (12.76) 5.34 (1.04)a,b 5.74 (0.84)c 5.64 (0.81)b,c 5.04 (1.13)a 5.81 (0.96)c 5.66 (0.90)b,c 5.45 (0.98)b,c 5.73 (0.97)c
Personaliz. (30.17) 6.02 (0.86)a,d 6.01 (0.81)a,d 6.25 (0.75)d 4.83 (1.22)c 6.08 (0.79)a,d 5.62 (0.85)b 5.76 (0.88)a,b 5.77 (0.80)a,b
Soc. Presence (220.62) 6.37 (0.72)a 3.75 (1.41)b 3.08 (1.22)f 3.71 (1.37)b 5.35 (1.05)d,e 5.65 (0.89)e 5.07 (1.26)d 4.94 (1.58)c
Anonymity (149.46) 1.98 (1.30)a 3.55 (1.52)c,d 4.20 (1.46)e,f 3.05 (1.32)b,c 2.87 (1.38)b 3.38 (1.44)b,c,d 3.76 (1.52)d,e 4.59 (1.63)f

COMMUNICATION MONOGRAPHS
Notes: Means and standard deviations were calculated using the final version of the PSACCS. All F-tests (7, 1476) are significant, p < .001. Significantly different means are indicated across rows by
different superscripts, p < .01.

15
16 J. FOX AND B. MCEWAN

affordances cited in the literature as crucial components of channel use, user experiences,
and communicative outcomes (e.g., boyd, 2011; Fulk et al., 1990; Treem & Leonardi,
2013). Further, the measure can be used to test or challenge fundamental assumptions
and predictions made by several channel theories and models.
Beyond the provision of the scale, our findings have some interesting implications. For
example, our results suggest that certain affordances outlined by previous literature may
be not be perceived distinctly by users. For example, the perceived affordance of synchro-
nicity is so tightly connected to perceptions of editability and social presence that synchro-
nicity failed to emerge as a distinct perceived affordance. Several models and theories
make attributions specific to asynchronicity, but many of these also relate its significance
to other affordances (e.g., the hyperpersonal model (Walther, 1996) suggests that asyn-
chronicity facilitates editability). Future research should explore whether asynchronicity
is distinctly meaningful, as the development of this scale is not sufficient to answer this
question completely. For scholars interested in pursuing work related to synchronicity,
our original items can be viewed in Table 1, although we must caution their use in con-
junction with other affordance measures.
Direct comparisons of channels on perceived social affordances also offered important
insights. Our findings indicate face-to-face communication has several perceived draw-
backs compared to other channels: it is less accessible, indicating communication is
more difficult to initiate; people feel less conversational control, so they may have difficulty
avoiding unpleasant interactions; and it is not persistent, so it may be hard to review what
was communicated in an important exchange. Early CMC researchers often privileged
face-to-face and considered it the preferred or desirable choice (Culnan & Markus,
1987; Daft et al., 1987). Although many CMC scholars have moved away from this
trend, face-to-face communication is often still heralded as superior by both scholars
and popular media (e.g., Turkle, 2015). Taking an affordance-based approach rather
than a channel-based approach may identify contexts in which face-to-face communi-
cation is not optimal and perhaps undesirable. It may also help overcome long-held
biases against certain channels, such as assuming that “lean” media like texting are inher-
ently low in bandwidth and thus insufficient for conveying emotion.
Measuring perceived social affordances could help identify explanatory mechanisms in
existing channel theories and findings. Perceived affordances could explain why or when
people switch channels in interpersonal relationships (modality switching; Ramirez &
Zhang, 2007) or why perceptions or effects of a message vary when delivered via different
channels (disclosure personalism; Bazarova, 2012). Measuring perceived affordances may
elucidate observed patterns, shifts, or outcomes of varied channel use in contexts such as
social support (High & Solomon, 2014) or everyday social interaction (Ruppel & Burke,
2015). Assessing perceived affordances could also help clarify boundary conditions for
theories that were developed before the emergence of certain technologies (Walther,
2009). For example, SIP suggests a slower process of information exchange during
relationship development via CMC. Is this pacing similar when people are interacting
through a social networking site like Facebook, where copious personal disclosures are
visible, previous interactions with others persist, and information can be obtained
through associated nodes?
In practice, when experimenters attempt to manipulate certain affordances, subscales of
this measure could also be employed as manipulation checks. They may also provide
 COMMUNICATION MONOGRAPHS 17

greater explanatory power and account for some observed variance in outcomes. Tao and
Bucy (2007) recommend treating the media attribute as the independent variable and its
associated psychological state as a mediator to illuminate our understanding of media
effects; our perceived affordances measure provides a consistent way to assess such
psychological states.
Because this measure allows researchers to assess perceptions and account for individ-
ual variability, another interesting direction for future research could be identifying factors
that may predict the perception of affordances. Previous experience or training may influ-
ence perceptions; for example, levels of technological education have been associated with
different expectations about online privacy (Buchanan et al., 2007). Other traits and indi-
vidual differences may also play a role. For example, attachment anxiety may predict dif-
fering perceptions of social presence, self-monitors may perceive editability differently, or
communication competence may predict how individuals perceive bandwidth. Situational
factors related to the context, content of communication, or relationship to the commu-
nicator may also influence perceptions. For example, a user may see texting as typically
private, but not if the recipient is a friend who gossips and is likely to show a text to others.
In this study, we assessed commonly cited affordances associated with human com-
munication across mediated and non-mediated channels. No single measure could poss-
ibly encapsulate every perceived affordance, however, so inevitably some are not included
here. Further, because this measure was intended to test existing theoretical claims and be
suitable for making direct comparisons to face-to-face communication (i.e., address
potential type I technology errors; Walther, 2013), all items needed to make sense in
the face-to-face context. Some affordances identified in other work that have potential
for communication (e.g., locatability; Schrock, 2015) are only sensibly applied to certain
channels. Further, to attain parsimony and cohesiveness, we did not assess a number of
affordances related more directly to human–computer interaction, such as the navigability
of an interface (Sundar, 2008).
Our study selected channels reflecting previous channel-based research (e.g., Kelly &
Keaten, 2007; Ledbetter, 2014; Ruppel & Burke, 2015), but was limited to face-to-face,
texting, phone, email, Facebook posts, instant messaging, video conferencing, and Snap-
chat. Future research could incorporate previously tested channels such as postal mail
or expand to newer channels such as multiplayer video games or collaborative virtual
environments.
In conclusion, this measure represents an attempt to advance the scientific understand-
ing of individuals’ experiences with communication channels. Given that affordances are
embedded in several communication theories and models, a systematic way to examine
perceived social affordances is long overdue. Focusing on perceived affordances will
help communication researchers continue to progress our understanding of social inter-
action that is viably situated in human experience rather than the ever-moving target of
technology.

Acknowledgements
We would like to thank participants of The Ohio State School of Communication’s Communication
Technology Symposium for their feedback during early conceptualizations of this project. We would
also like to thank the Editor and three anonymous reviewers for their insights and helpful comments.
18 J. FOX AND B. MCEWAN

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This research was funded in part by The Ohio State University School of Communication Mattox
Award granted to the first author.

References
Bazarova, N. N. (2012). Public intimacy: Disclosure interpretation and social judgments on
Facebook. Journal of Communication, 62, 815–832. doi:10.1111/j.1460-2466.2012.01664.x
Boster, F. J., Kotowski, M. R., Andrews, K. R., & Serota, K. (2011). Identifying influence:
Development and validation of the connectivity, persuasiveness, and maven scales. Journal of
Communication, 61, 178–196. doi:10.1111/j.1460-2466.2010.01531.x
boyd, d. (2011). Social network sites as networked publics: Affordances, dynamics, and impli-
cations. In Z. Papacharissi (Ed.), A networked self: Identity, community, and culture on social
network sites (pp. 39–58). New York, NY: Routledge.
boyd, d. m., & Ellison, N. B. (2007). Social network sites: Definition, history, and scholarship.
Journal of Computer-Mediated Communication, 13, 210–230. doi:10.1111/j/1083-6101.2007.
00393.x
Bradner, E., Kellogg, W. A., & Erickson, T. (1999). The adoption and use of ‘Babble’: A field study of
chat in the workplace. In Ecscw’99 (pp. 139–158). Copenhagen: Springer.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S.
Long (Eds.), Testing structural equation models (pp. 136–162). Beverly Hills, CA: Sage.
Buchanan, T., Paine, C., Joinson, A. N., & Reips, U. D. (2007). Development of measures of online
privacy concern and protection for use on the Internet. Journal of the American Society for
Information Science and Technology, 58, 157–165. doi:10.1002/asi.20459
Carlson, J. R., & Zmud, R. W. (1999). Channel expansion theory and the experiential nature of
media richness perceptions. Academy of Management Journal, 42(2), 153–170. doi:10.2307/
257090
Clark, H. H., & Brennan, S. E. (1991). Grounding in communication. In L. B. Resnick, J. M. Levine,
& S. D. Teasley (Eds.), Perspectives on socially shared cognition (pp. 127–149). Washington, DC:
American Psychological Association.
Culnan, M. J., & Markus, M. L. (1987). Information technologies. In F. M. Jablin, L. L. Putnam, K.
H. Roberts, & L. W. Porter (Eds.), Handbook of organizational communication (pp. 420–443).
Newbury Park, CA: Sage.
Daft, R. L., Lengel, R. H., & Trevino, L. K. (1987). Message equivocality, media selection, and
manager performance: Implications for information systems. MIS Quarterly, 11, 355–366.
doi:10.2307/248682
DeAndrea, D. C., Van Der Heide, B., Vendemia, M. A., & Vang, M. H. (in press). How people evalu-
ate online reviews. Communication Research. doi:10.1177/0093650215573862
DeVellis, R. F. (2017). Scale development: Theory and applications. Thousand Oaks, CA: Sage.
Ellison, N. B., & Vitak, J. (2015). Social network site affordances and their relationship to social
capital processes. In S. S. Sundar (Ed.), The handbook of the psychology of communication tech-
nology (pp. 205–227). New York, NY: Wiley & Sons.
Eveland, W. P. (2003). A “mix of attributes” approach to the study of media effects and new com-
munication technologies. Journal of Communication, 53, 395–410. doi:10.1111/j.1460-2466.
2003.tb02598.x
Feaster, J. C. (2010). Expanding the impression management model of communication channels:
An information control scale. Journal of Computer-Mediated Communication, 16, 115–138.
doi:10.1111/j.1083-6101.2010.01535.x
 COMMUNICATION MONOGRAPHS 19

Fox, J., & Moreland, J. J. (2015). The dark side of social networking sites: An exploration of the rela-
tional and psychological stressors associated with Facebook use and affordances. Computers in
Human Behavior, 45, 168–176. doi:10.1016/j.chb.2014.11.083
Fulk, J., Schmitz, J., & Steinfield, C. W. (1990). A social influence model of technology use. In J. Fulk
& C. W. Steinfield (Eds.), Organizations and communication technology (pp. 117–140). Newbury
Park, CA: Sage.
Gaver, W. W. (1991). Technology affordances. Proceedings of the SIGCHI Conference on Human
Factors, 79–54. doi:10.1145/108844.108856
Gibson, J. J. (1979). The ecological approach to visual perception. Boston, MA: Houghlin Mifflin.
Haythornthwaite, C. (2002). Strong, weak, and latent ties and the impact of new media. The
Information Society, 18, 385–401. doi:10.1080/01972240290108195
High, A. C., & Solomon, D. H. (2014). Communication channel, sex, and the immediate and longi-
tudinal outcomes of verbal person-centered support. Communication Monographs, 81, 439–468.
doi:10.1080/03637751.2014.933245
Hogan, B. J. (2009). Networking in everyday life (Doctoral dissertation). University of Toronto.
Retrieved from http://individual.utoronto.ca/berniehogan/Hogan_NIEL_10-29-2008_FINAL.
pdf
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling: A
Multidisciplinary Journal, 6, 1–55. doi:10.1080/10705519909540118
Hutchby, I., & Barnett, S. (2005). Aspects of the sequential organization of mobile phone conversa-
tion. Discourse Studies, 7, 147–171. doi:10.1177/1461445605050364
Kelly, L., & Keaten, J. A. (2007). Development of the affect for communication channels scale.
Journal of Communication, 57, 349–365. doi:10.1111/j.1460-2466.2007.00346.x
Kline, R. B. (2016). Principles and practice of structural equation modeling (4th ed.). New York, NY:
Guilford.
Lea, M., & Spears, R. (1991). Computer-mediated communication, deindividuation, and group
decision-making. International Journal of Man-Machine Studies, 34, 283–301. doi:10.1016/
0020-7373(91)90045-9
Ledbetter, A. M. (2009). Measuring online communication attitude: Instrument development and
validation. Communication Monographs, 76, 463–486. doi:10.1080/03637750903300262
Ledbetter, A. M. (2014). Online communication attitude similarity in romantic dyads: Predicting
couples’ frequency of e-mail, instant messaging, and social networking site communication.
Communication Quarterly, 62, 233–252. doi:10.1080/01463373.2014.890120
Lee, E. J., & Oh, S. Y. (2012). To personalize or depersonalize? When and how politicians’ person-
alized tweets affect the public’s reactions. Journal of Communication, 62, 932–949. doi:10.1111/j.
1460-2466.2012.01681.x
Lombard, M., Ditton, T. B., & Weinstein, L. (2009, November). Measuring (tele)presence: The
Temple Presence Inventory. Presented at the Twelfth International Workshop on Presence, Los
Angeles, CA.
Marsh, H. W., Hau, K-T., & Wen, Z. (2004). In search of golden rules: Comment on hypothesis-
testing approaches to setting cutoff values for fit indexes and dangers in overgeneralizing Hu
and Bentler’s (1999) findings. Structural Equation Modeling: A Multidisciplinary Journal, 11,
320–341. doi:10.1207/s15328007sem1103_2
McOmber, J. B. (1999). Technological autonomy and three definitions of technology. Journal of
Communication, 49, 137–153. doi:10.1111/j.1460-2466.1999.tb02809.x
Meyrowitz, J. (2009). Medium theory: An alternative to the dominant paradigm of media effects. In
R. L. Nabi & M. B. Oliver (Eds.), The Sage handbook of media processes and effects (pp. 517–530).
Thousand Oaks, CA: Sage.
Nass, C., & Mason, L. (1990). On the study of technology and task: A variable-based approach. In J.
Fulk & C. Steinfield (Eds.), Organizations and communication technology (pp. 46–67). Newbury
Park, CA: Sage.
Norman, D. A. (1990). The design of everyday things. New York, NY: Doubleday.
20 J. FOX AND B. MCEWAN

O’Sullivan, P. B. (2000). What you don’t know won’t hurt me: Impression management functions of
communication channels in relationships. Human Communication Research, 26, 403–431.
doi:10.1111/j.1468-2958.2000.tb00763.x
Parks, M. R. (2008). Characterizing the communicative affordances of MySpace: A place for friends
or a friendless place? Paper presented at the Annual Conference of the International
Communication Association, Montreal.
Pearce, K. E., & Rice, R. E. (2013). Digital divides from access to activities: Comparing mobile and
personal computer Internet users. Journal of Communication, 63, 721–744. doi:10.1111/jcom.
12045
Pew Research Center. (2015). Technology device ownership 2015. Washington, DC: Author.
Rains, S. A. (2007). The impact of anonymity on perceptions of source credibility and influence in
computer-mediated group communication: A test of two competing hypotheses.
Communication Research, 34, 100–125. doi:10.1177/0093650206296084
Ramirez Jr., A., Dimmick, J., Feaster, J., & Lin, S. F. (2008). Revisiting interpersonal media compe-
tition: The gratification niches of instant messaging, e-mail, and the telephone. Communication
Research, 35, 529–547. doi:10.1177/0093650208315979
Ramirez Jr., A., & Zhang, S. (2007). When online meets offline: The effect of modality switching on
relational communication. Communication Monographs, 74, 287–310. doi:10.1080/
03637750701543493
Reid, F. J. M., & Reid, D. J. (2010). The expressive and conversational affordances of mobile messa-
ging. Behaviour & Information Technology, 29, 3–22. doi:10.1080/014492907014970
Resnick, P. (2001). Beyond bowling together: Sociotechnical capital. In J. Carroll (Ed.), HCI in the
new millennium (pp. 247–272). New York, NY: Addison-Wesley.
Rice, R. E. (1987). Computer-mediated communication and organizational innovation. Journal of
Communication, 37(4), 65–94. doi:10.1111/j.1460-2466.1987.tb01009.x
Rice, R. E., & Steinfield, C. (1994). Experiences with new forms of organizational communication
via electronic mail and voice messaging. In J. H. E. Andriesson & R. A. Roe (Eds.), Telematics and
work (pp. 109–132). Hillsdale, NJ: Erlbaum.
Ritter, B. A. (2014). Deviant behavior in computer-mediated communication: Development and
validation of a measure of cybersexual harassment. Journal of Computer-Mediated
Communication, 19, 197–214. doi:10.1111/jcc4.12039
Rogers, E. M. (1962). Diffusion of innovations. New York, NY: Free Press.
Ruppel, E. K., & Burke, T. J. (2015). Complementary channel use and the role of social competence.
Journal of Computer-Mediated Communication, 20, 37–51. doi:10.1111/jcc4.12091
Schmitz, J., & Fulk, J. (1991). Organizational colleagues, media richness, and electronic mail: A test
of the social influence model of technology use. Communication Research, 18, 487–523. doi:10.
1177/009365091018004003
Schouten, A., Valkenburg, P. M., & Peter, J. (2007). Precursors and underlying processes of adoles-
cents’ online self-disclosure: Developing and testing an “Internet-attribute-perception” model.
Media Psychology, 10, 292–315. doi:10.1080/15213260701375686
Schrock, A. R. (2015). Communicative affordances of mobile media: Portability, availability, locat-
ability, and multimediality. International Journal of Communication, 9, 1229–1246. Retrieved
from http://ijoc.org/index.php/ijoc/article/view/3288
Short, J., Williams, E., & Christie, B. (1976). The social psychology of telecommunications. New York,
NY: Wiley.
Spitzberg, B. H. (2006). Preliminary development of a model and measure of computer-mediated
communication (CMC) competence. Journal of Computer-Mediated Communication, 11, 629–
666. doi:10.1111/j.1083-6101.2006.00030.x
Steinfield, C. W. (1986). Computer-mediated communication in an organizational setting:
Explaining task-related and socioemotional uses. In M. L. McLaughlin (Ed.), Communication
yearbook (Vol. 9, pp. 777–804). Newbury Park, CA: Sage.
Stoycheff, E. (2016). Under surveillance: Examining Facebook’s spiral of silence effects in the wake
of NSA internet monitoring. Journalism & Mass Communication Quarterly, 93, 296–311. doi:10.
1177/1077699016630255
 COMMUNICATION MONOGRAPHS 21

Suh, A., & Wagner, C. (2013, January). Factors affecting individual flaming in virtual communities.
In Proceedings of the 46th Hawaii International Conference on System Sciences (HICSS)
(pp. 3282–3291). IEEE. doi:10.1109/HICSS.2013.230
Sundar, S. S. (2008). The MAIN model: A heuristic approach to understanding technology effects
on credibility. In M. J. Metzger & A. J. Flanagin (Eds.), Digital media, youth, and credibility (pp.
72–100). Cambridge, MA: MIT Press.
Sundar, S. S., & Bellur, S. (2011). Concept explication in the Internet age: The case of interactivity.
In E. P. Bucy & R. L. Holbert (Eds.), Sourcebook for political communication research: Methods,
measures, and analytical techniques (pp. 485–500). New York: Routledge.
Sundar, S. S., & Limperos, A. M. (2013). Uses and grats 2.0: New gratifications for new media.
Journal of Broadcasting & Electronic Media, 57, 504–525. doi:10.1080/08838151.2013.845827
Tanis, M., & Postmes, T. (2003). Social cues and impression formation in CMC. Journal of
Communication, 53, 676–693. doi:10.1111/j.1460-2466.2003.tb02917.x
Tao, C. C., & Bucy, E. P. (2007). Conceptualizing media stimuli in experimental research:
Psychological versus attribute-based definitions. Human Communication Research, 33, 397–
426. doi:10.1111/j.1468-2958.2007.00305.x
Toma, C. L., Hancock, J. T., & Ellison, N. B. (2008). Separating fact from fiction: An examination of
deceptive self-presentation in online dating profiles. Personality and Social Psychology Bulletin,
34, 1023–1036. doi:10.1177/0146167208318067
Treem, J., & Leonardi, P. (2013). Social media use in organizations: Exploring the affordances of
visibility, editability, persistence, and association. In C. T. Salmon (Ed.), Communication year-
book (vol. 38, pp. 143–189). New York, NY: Routledge.
Turkle, S. (2015). Reclaiming conversation: The power of talk in the digital age. New York, NY:
Penguin Press.
Walther, J. B. (1994). Anticipated ongoing interaction versus channel effects on relational com-
munication in computer-mediated interaction. Human Communication Research, 20, 473–
501. doi:10.1111/j.1468-2958.1994.tb00332.x
Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal and hyper-
personal interaction. Communication Research, 23, 3–43. doi:10.1177/009365096023001001
Walther, J. B. (2009). Computer-mediated communication. In C. R. Berger, M. E. Roloff, & D. R.
Roskos-Ewoldsen (Eds.), Handbook of communication science (2nd ed., pp. 489–505). Newbury
Park, CA: Sage.
Walther, J. B. (2013). Affordances, effects, and technology errors. In C. T. Salmon (Ed.),
Communication yearbook (vol. 38, pp. 190–193). New York, NY: Routledge.
Walther, J. B., Gay, G., & Hancock, J. T. (2005). How do communication and technology research-
ers study the internet? Journal of Communication, 55, 632–657. doi:10.1111/j.1460-2466.2005.
tb02688.x
Walther, J. B., & Parks, M. R. (2002). Cues filtered out, cues filtered in: Computer-mediated com-
munication and relationships. In M. L. Knapp & J. A. Daly (Eds.), Handbook of interpersonal
communication (3rd ed., pp. 529–563). Thousand Oaks, CA: Sage.
Wellman, B., Quan-Haase, A., Boase, J., Chen, W., Hampton, K., Díaz, I., & Miyata, K. (2003). The
social affordances of the Internet for networked individualism. Journal of Computer-Mediated
Communication, 8(3). doi:10.1111/j.1083-6101.2003.tb00216.x
Williams, F., & Rice, R. E. (1983). Communication research and the new media technologies. In R.
N. Bostrom (Ed.), Communication yearbook (Vol. 7, pp. 200–224). Beverly Hills, CA: Sage.