Looking Within: An Examination, Combination, and Extension of Within-Person Methods Across Multiple Levels of Analysis

Download as pdf or txt
Download as pdf or txt
You are on page 1of 23

Chapter 13

LOOKING WITHIN: AN EXAMINATION,


COMBINATION, AND EXTENSION OF
WITHIN-PERSON METHODS ACROSS
MULTIPLE LEVELS OF ANALYSIS
Daniel J. Beal and Allison S. Gabriel
Copyright American Psychological Association. Not for further distribution.

Over the past 20 years, there has been a rapid (Blau, 1964; Cropanzano & Mitchell, 2005) and
increase in intraindividual studies in the social theories highlighting dynamic exchanges between
and behavioral sciences. Indeed, recent searches interaction partners (Andersson & Pearson, 1999;
of the literature for studies utilizing a suite of Groth & Grandey, 2012)—all have dynamic under­
intraindividual methods—from daily diaries to pinnings, meaning that, in general, person-level
experience sampling methodology (ESM)—indicate assessments of constructs implied by these theories
that over 300 studies have been published using may be missing the empirical mark.
this methodological approach, with more than In light of the ease of implementation of and the
half of these studies emerging in just the last theoretical necessity for intraindividual research,
6 years alone (Clarivate Analytics, 2018). In part, scholars have implemented a variety of method­
the increase of intraindividual research has come ological approaches, including daily diaries, ESM,
about due to the fact that the methods are becoming and, more recently, continuous rating assessments
increasingly more user-friendly (e.g., ESM and diary (CRAs). The goal of each approach is to provide
studies can be implemented using online survey insight into day-to-day, event-to-event, and even
platforms; smartphone apps are being created) and moment-to-moment fluctuations in attitudes,
more analytic resources and approaches are being emotions, and experiences of individuals. However,
developed to help scholars analyze the complex although each approach has its merits, little work has
data structures that intraindividual methods yield considered the ways in which the approaches can be
(e.g., multilevel moderated mediation with Mplus used in tandem; that is, while reviews of the various
and R; discontinuous growth modeling; lagged intraindividual approaches exist (e.g., Beal, 2015;
panel analyses). Additionally, the increase in this Gabriel, Diefendorff, Bennett, & Sloan, 2017; Ohly,
approach has coincided with growing awareness Sonnentag, Niessen, & Zapf, 2010), they are often
that many of the theories that we use in the social viewed as occurring in isolation from one another.
sciences have dynamic components that need to be This is interesting, given that scholars have argued
properly addressed. For example, within management that multilevel research should integrate what is
and applied psychology, severally commonly applied occurring across various levels of analysis. Bliese,
theories—from affective events theory (Weiss & Chan, and Ployhart (2007; see also Chen, Bliese,
Cropanzano, 1996) and episodic views of performance & Mathieu, 2005) encouraged scholars to focus on
(Beal, Weiss, Barros, & MacDermid, 2005; Morgeson, homology, or the consideration of how effects may
Mitchell, & Liu, 2015) to social exchange theory change as researchers move across levels. Although

http://dx.doi.org/10.1037/0000115-014
The Handbook of Multilevel Theory, Measurement, and Analysis, S. E. Humphrey and J. M. LeBreton (Editors-in-Chief)
Copyright © 2019 by the American Psychological Association. All rights reserved.

305
Beal and Gabriel

this has led some scholars to consider how effects environment, (b) assessing constructs and/or
change at the person- and event-level of analysis events close to their actual occurrence capturing
(e.g., Gabriel, Diefendorff, Chandler, Moran, & the immediacy of the experience, and (c) ensuring
Greguras, 2014), we contend that this consideration representative sampling of individuals’ experiences.
likely applies to moving across various temporal The first point captures events in the environment as
levels embedded within various intraindividual opposed to being constrained by laboratory contexts
techniques. That is, as researchers move across lower (although we note that one method, CRA, can occur
levels of analysis, important insights about temporal in experimental settings); this helps enhance the
patterns can be gained, in addition to knowledge about ecological validity of the constructs being captured.
how constructs may emerge at various time points. The second point helps in eliminating retrospective
Our goal in the current chapter is to highlight biases (Robinson & Clore, 2002) by capturing
how methods to examine intraindividual phenomena experiences closer to the time that they are actually
can be used separately or in tandem to achieve this occurring, as opposed to asking participants to recall
Copyright American Psychological Association. Not for further distribution.

understanding. This chapter first explores the space or mentally aggregate over extended time frames. The
surrounding the within-person level of analysis as last point focuses on the fact that, when assessing
it relates to different descriptions of immediate, events and/or constructs repeatedly over several
lived experiences of individuals (e.g., Weiss & hours, days, and weeks, researchers can help ensure
Rupp, 2011). Then, we describe an integrative, that they are getting a fairly accurate representation of
programmatic method for linking these approaches. the lived experience of individuals (Beal, Trougakos,
Finally, we move these methods to a higher level Weiss, & Dalal, 2013; Weiss & Rupp, 2011). Thus,
of analysis, describing how insights gained at the although there are nuances in the intraindividual
within-person level can inform emergent models approaches outlined below, all five aim at achieving
at the individual level of analysis and identifying these three points. Instead, the primary differences
some of the potential hurdles such models might across the five approaches is the time referent, or level
encounter. Importantly, the purpose of this chapter of aggregation, being considered.
is not to just provide an overview—or how-to The daily ESM approach involves using ESM
guide—for conducting research with intraindividual methods to capture constructs that are recalled
approaches (for reviews achieving this objective across an entire day. As such, for scholars who are
to gain familiarity with the approaches, see Beal, perhaps unclear of the pacing of the construct(s)
2015; Beal & Weiss, 2003; Dimotakis, Ilies, & being studied (i.e., it is unclear how quickly or how
Judge, 2013; Fisher & To, 2012; Gabriel et al., long it takes a process to unfold), having participants
2017; Ohly et al., 2010). Rather, we aim to present reflect upon the entire workday can prove fruitful.
an overview of the ways in which scholars using With this approach, organizational researchers might
intraindividual methods can implement these ask employees to recall their level of performance
approaches collaboratively to delve deeper into on various tasks throughout the day (e.g., Gabriel,
theoretical questions of interest. Thus, having a Diefendorff, & Erickson, 2011), the types and level
preliminary understanding of what ESM and CRA of various emotion-regulation strategies used during
methods represent will prove fruitful for interpreting interactions (e.g., Judge, Woolf, & Hurst, 2009;
the current chapter. Wagner, Barnes, & Scott, 2014), the various demands
experienced over the course of the day (e.g., Rodell &
Judge, 2009), or ratings of the behaviors exhibited by
TEMPORAL LEVELS OF FIVE
others at work, including coworkers and supervisors
WITHIN-PERSON METHODS
(e.g., Barnes, Lucianetti, Bhave, & Christian, 2015;
Across the five approaches to within-person designs Matta, Erol-Korkmaz, Johnson, & Biçaksiz, 2014). In
we describe below, there are three unify­ing character­ studying social interactions outside of the workplace,
istics highlighted by Beal (2015): (a) measuring spouses could be asked to recall their interactions
an event or experience occurring in the natural with each other over the course of a day, reflecting

306
Looking Within

on partner responsiveness and intimacy during Unlike the daily approach that focuses on
discussions with each other (e.g., Laurenceau, aggregations of experiences over an entire day,
Barrett, & Rovine, 2005), or children being the episodic approach shifts down a temporal
asked to report on peer harassment that occurred level, assessing briefer periods of time and further
while at school on a given day (e.g., Nishina & reducing the amount of mental aggregation that study
Juvonen, 2005). participants are required to make. For example,
As an example from organizational research, researchers who use ESM with a signal-contingent
Yang and Diefendorff (2009) assessed the daily approach—that is, an approach in which participants
relationships between counterproductive work respond to a survey when they are signaled to do so
behaviors targeted toward the organization and (e.g., by a survey being e-mailed at a planned and/or
individuals, negative emotions, and various situa­ random time)—may ask study participants to
tional antecedents, such as ambiguity and both respond to survey items since the last time they
supervisor and coworker interpersonal (in)justice. were surveyed (e.g., “since the last beep”), over a
Copyright American Psychological Association. Not for further distribution.

As part of their ESM procedure, participants were specific period of time (e.g., “over the last 2 hours”),
administered a survey at the end of each workday or in reference to a specific event (e.g., “during the
online at 4:00 p.m. and were asked to complete the last social interaction”; “during the last task you
survey prior to leaving work. Each item that was completed”). Conversely, researchers can also use
part of the ESM survey was adapted to be assessed an event-contingent approach, which asks study
on a daily basis. For instance, regarding customer participants to complete ESM surveys only when a
interpersonal injustice, participants responded to particular event happens to them (e.g., when women
items such as: “Regarding your customers’ behavior are in a social interaction in which they decided to
today, to what extent did your customers treat you disclose or hide their pregnancy [Jones, 2017]; when
with respect?” (Yang & Diefendorff, 2009, p. 271; an e-mail is received after hours [Butts, Becker, &
italics added for emphasis). As another example of Boswell, 2015]; when experiencing a particular
a daily approach, in a dyadic ESM study focusing issue and/or occurrence associated with an eating
on performance and work engagement, Bakker or psychological disorder [Berg et al., 2013]). This
and Xanthopoulou (2009) had coworker pairs latter approach places a heavy burden on research
each complete a survey at the end of the workday, participants, requiring them to remember to complete
with items focusing on mental aggregations over a survey shortly after the occurrence of the event of
the entire day. For instance, items focusing on interest to the researchers.
engagement reflected how engaged employees Theoretically, episodes in this approach refer
were over the entire workday (e.g., “Today, I felt to subjective divisions of an individual’s daily
strong and vigorous while working”), as were experience. Beal and colleagues (Beal et al., 2005;
performance items (e.g., “Today, I fulfilled all the Beal & Weiss, 2003) have discussed this idea
requirements of my job”; Bakker & Xanthopoulou, extensively; to summarize, an episode refers to the
2009, p. 1565; italics added for emphasis). Of natural manner in which individuals chunk their
course, the daily ESM approach implicitly allows ongoing experiences into coherent, goal-relevant
for a great likelihood of recall bias to enter into segments. Typically, these segments allow for key
the evaluations being made compared to the features of the episode to be encoded more effectively
other approaches described in detail below (e.g., and help impart meaning to what would otherwise be
reflecting on engagement over the entire workday a continuous stream of information (Barker, 1968;
makes it unclear how engagement in the morning Swallow, Zacks, & Abrams, 2009). Ideally, survey
and afternoon is being mentally weighted or items using this approach would be framed around
recalled). Nonetheless, this approach provides specific, identifiable episodes in an individual’s day.
researchers with the opportunity to understand This means that researchers need to think carefully
how phenomena of interest change from one day about issues concerning the identifiability of
to the next. particular types of episodes and how quickly the

307
Beal and Gabriel

phenomenon of interest will unfold in the time if participants did not experience such an event,
frame being assessed (Gabriel et al., 2017; Monge, they were further instructed to not provide any
1990). For instance, if researchers are examining data. All survey items were asked in referent to the
mentor–mentee interactions and have fairly short e-mail event; for instance, participants assessed the
time frames being assessed with their episodic affective tone of the e-mail, how long it took them to
approach, it is possible that an interaction between respond to the e-mail, and how they felt affectively
a mentor and his or her protégé may not even after reading the e-mail. As a final example, Beal and
occur during the specified time frame, making the colleagues (2013) sought to break down an episode
measurement inaccurate and invalid. Importantly, into smaller chunks, asking a sample of restaurant
this issue of time is one that affects all ESM and employees to complete surveys before and after a
intraindividual approaches and has been the focus of mealtime rush.
theorizing by Monge (1990) and others (e.g., George A blend of the daily and episodic ESM approaches
& Jones, 2000; Mitchell & James, 2001; Ployhart & would be the day reconstruction method (DRM;
Copyright American Psychological Association. Not for further distribution.

Vandenberg, 2010), who have lamented that scholars Kahneman, Krueger, Schkade, Schwarz, & Stone,
who wish to assess research questions associated 2004) or a variant of that approach. According
with dynamics and time need to carefully select to Kahneman et al. (2004), this method takes
the time frame being referenced in the construct components of episodic ESM in that it has partici­
assessments (as an example, see Methot, Lepak, pants think about particular salient events. However,
Shipp, & Boswell, 2017). participants in a DRM study, rather than being
Several examples of an episodic approach to ESM surveyed after the event or several times throughout
exist. Within the organizational sciences, Benedetti, the day, are asked to write down and rebuild specific
Diefendorff, Gabriel, and Chandler (2015) were events that happened throughout the workday at
interested in how employees experienced different a later time, which minimizes the interruptions
types of motivation at work. Specifically, Benedetti associated with ESM. For instance, participants
et al. (2015) assessed participants’ motivation for would first write down a description of an event
pursuing their last work-related task five times a day (e.g., what occurred or who was involved) and
for 10 workdays; these surveys were administered recall feelings experienced during the event; as
via PalmPilots using the Purdue Momentary stated by Kahneman et al., “the goal is to provide
Assessment Tool (Weiss, Beal, Lucy, & MacDermid, an accurate picture of the experience associated
2004) to randomly signal participants throughout with activities [e.g., commuting] and circumstances
the workday at semirandom times that were at [e.g., a job with time pressure]” (p. 1776). Thus,
least 75 minutes apart. The motivation items were a well-executed DRM study should help people
prefaced with instructions asking participants to relive what occurred to them, reducing biases
think about the last work-related task they had typically associated with recalling information. As
completed. Participants then rated the extent the DRM asks for a more specific reconstruction of
to which the task was pursued because it was daily experiences, it in many ways blends daily and
enjoyable and interesting (i.e., intrinsic motivation) episodic approaches. The key distinction between
versus it being because their supervisor asked them the DRM as originally described by Kahneman and
to complete it or the situation demanded it (i.e., colleagues and any ESM variant is that only a single
extrinsic motivation). These instructions helped day is reconstructed for each participant.
participants reflect back on one salient performance Despite the necessarily limited range of experiences
episode. In another recent example, Butts et al. captured per person, Kahneman et al. (2004)
(2015) asked participants to complete surveys in suggested that the DRM still provides a representative
the evening after they received work-related e-mail, range of situations and experiences by allowing for
stating: “Complete the survey based upon the most larger samples of individuals with less burden of
recent electronic communication received from participation. Finally, although not recommended
work after working hours” (p. 773). Importantly, by Kahneman et al., it is possible to blend the DRM

308
Looking Within

with daily ESM. This approach would—at the end p. 41, italics added for emphasis). Other researchers
of each day of the multiday study—ask people to have taken this approach, with Gabriel et al. (2014)
reconstruct their experiences by walking through assessing perceptions of person–organization and
and rating each episode sequentially. The goal with person–job fit perceptions at the survey signal, and
this method would be to obtain a more precise Beal and colleagues (2013) assessing the extent to
and continuous recall and judgment of a day’s which restaurant employees currently felt fatigued.
events relative to simply asking people to mentally Whereas diaries and ESM approaches often are
summarize, as is typically the case in daily ESM focused on mentally aggregating experiences over
studies. weeks, days, hours, and/or specific events (e.g., Beal,
As an example of this hybrid approach, Bakker, 2015; Ohly et al., 2010), our final approach—
Demerouti, Oerlemans, and Sonnentag (2013) continuous rating assessments (CRAs)—can be
examined workaholism and recovery experiences. viewed as the continuous approach, focusing on
Bakker et al. had participants for 9 workdays each capturing unfolding dynamics within a specific
Copyright American Psychological Association. Not for further distribution.

evening before going to bed select activities from a list event (Gabriel, Diefendorff, Bennett, & Sloan,
that they engaged in during the evening, indicating 2017; Ruef & Levenson, 2007). More specifically,
how long they spent on each activity and their theories and constructs assessed with diaries and
happiness associated with each activity. Likewise, ESM assessments—such as emotions (e.g., Beal,
Oerlemans and Bakker (2014) had participants for Trougakos, Weiss, & Green, 2006; Scott & Barnes,
2 workweeks complete a survey in the morning in 2011), social exchange relationships (e.g., Rosen,
which they reconstructed the time they spent on Koopman, Gabriel, & Johnson, 2016), job stressors
off-job activities after they left work the previous day. and demands (e.g., Dimotakis, Scott, & Koopman,
Participants were asked to rebuild their postwork 2011; Sonnentag & Zijlstra, 2006), social support
time in chronological order and then report on their (Eisenberger, Taylor, Gable, Hilmert, & Lieberman,
current level of recovery. As noted by Oerlemans 2007; Siewert, Antoniw, Kubiak, & Weber, 2011),
and Bakker, using DRM helped facilitate “access to and marital communication and intimacy (Laurenceau
encoded momentary experiences” that occurred the et al., 2005)—often are presumed to ebb and flow
prior evening, reducing recall biases (p. 306). over the course of a day or an event. Stated differently,
If a random sampling of an individual’s in a given interaction between a person and his
day-to-day experiences are desired rather than or her environment, he or she may have moment-
experiences surrounding a particular event or time to-moment fluctuations in mood, perceptions,
frame, another measurement approach can be used, and/or feelings of stress contingent upon changes
which we refer to as momentary ESM. This approach, continuously experienced over time. Thus, although
which reflects the original form of experience ESM methods capture change that occurs between
sampling methods (Csikszentmihalyi, Larson, & events or day to day, these approaches miss the
Prescott, 1977), focuses on assessing key constructs changes and trajectories of constructs that occur
“in the moment” as participants respond to survey within an event as it dynamically unfolds over time.
signals. For instance, using the signal-contingent As reviewed by Gabriel et al. (2017; see also
approach, when participants are sent an ESM survey, Ruef & Levenson, 2007), CRA is a within-person
they may be asked questions in which they respond methodological approach that can better capture
about emotions and cognitions “right now” or “at the changes in constructs within a salient event. To do
present moment.” For instance, in the Benedetti this, CRA captures moment-to-moment variations
et al. (2015) study noted above, the researchers were by asking individuals to continuously assess their
also interested in assessing job satisfaction and vitality feelings and cognitions as an event is unfolding;
at the present moment. Accordingly, items such as this event can either be happening to the participant
“at this moment, I am satisfied with my job” and for the first time during the ratings (i.e., online
“at this moment, I feel alive and vital” were asked when or in vivo ratings) or during a cued recall of an
participants were signaled (Benedetti et al., 2015, event. These ratings are collected via a handheld or

309
Beal and Gabriel

computerized rating dial that allows participants to the call and provided separate continuous ratings
adjust their ratings as much or as little as possible using a computer program of their felt emotions
in order to accurately reflect their current level of a during the call, as well as their use of two emotion-
given construct. Moreover, as evidenced in research regulation strategies—surface acting and deep acting
conducted in social psychology (e.g., Mauss, (Grandey, 2000)—in a counterbalanced order.
Levenson, McCarter, Wilhelm, & Gross, 2005; Ratings were collected every 200 milliseconds and
Mauss et al., 2011), self-reported CRAs can also aggregated to create 1-second averages for analyses.
be combined with other dynamic assessments such In addition, third-party raters who were blind
as physiological markers of arousal (e.g., galvanic to the study hypotheses and condition listened
skin response, heart rate, eye tracking, facial coding back to the call and provided continuous ratings
technology). Importantly, CRA can be a valuable of how positive or negative the vocal tone of the
tool for better addressing time-related issues over participant was. As such, CRA not only captures
relatively short periods of time, such as better emotions and cognitions associated with an event,
Copyright American Psychological Association. Not for further distribution.

understanding the rate of change of constructs, but dynamic assessments of performance-related


as well as the permanence of change, magnitude constructs as well.
of change, and the lag of change that occurs
(see Monge, 1990, for a review). Conceptual Similarities and Differences
Original work on CRA emerged in marital Among the Five Methods
research (e.g., Gottman & Levenson, 1985; Levenson Ultimately, our goal in this chapter is to initially
& Gottman, 1983; see Ruef & Levenson, 2007, for a outline the ways various within-person data
review). Within this line of work, researchers wanted collection methods connect (see also Gabriel et al.,
to allow marital interactions to naturally unfold, 2017). Doing so will provide a more complete
versus stopping the natural conversation at random picture of how a particular experience or set of
points to collect spouses’ assessments of affect and experiences unfold, from the most immediate
other psychological constructs (i.e., the “talk-table” reactions and fluctuations to the more aggregate
approach; Gottman et al., 1976). As highlighted progression of daily life. However, before we link
by Ruef and Levenson (2007), switching to CRA these different methods together, or discuss how
in marital research led to a much less intrusive one might go about doing so, it will be helpful
measurement approach and enhanced external to consider some of the central similarities and
validity given that there were fewer interruptions differences of the five methods that largely relate
being made in the natural conversation. In the to the timing of the approaches in assessing
organizational sciences, there are two studies that phenomena of interest. As described above, each
have used CRA. Naidoo and Lord (2008) used online approach to exploring within-person variability
CRA to assess perceptions of leader charisma. In focuses on a different timing or pacing of relevant
this study, Naidoo and Lord manipulated the level constructs, with the fastest pacing occurring with
of imagery being expressed in audio recordings of CRA, followed by momentary ESM, episodic ESM,
actors reciting the inaugural address of U.S. president DRM, and daily approaches. However, beyond
Franklin Delano Roosevelt. As participants were this surface-level difference are more fundamental
listening to the presidential speech, they moved a elements of what each method tries to capture.
mouse on a computer connected to a CRA scale, We have summarized these elements in Table 13.1
where 0 corresponded to not at all charismatic and and framed them around three common research
100 corresponded to extremely charismatic. In goals that within-person researchers strive to
another example, Gabriel and Diefendorff (2015) achieve: (a) capturing a representative sample of an
put participants in a call-center simulation in which individual’s varying experiences, (b) mapping patterns
they were exposed to a hostile customer caller. of change in these experiences, and (c) linking one
Using cued-recall CRA, after going through the stream of experiences to another. We split the latter
difficult customer call, participants listened back to category into a goal to link intraindividual experiences

310
Looking Within

TABLE 13.1

Capacity of Five Categories of Within-Person Methods to Achieve a Set of Conceptual Within-Person


Research Goals

Within-person research goal CRA Momentary ESM Episodic ESM DRM Daily ESM
Representative sample of an individual’s experiences − +++ + + +
Comprehensive account of dynamics +++ + ++ ++ ++
Capacity to link sequential experiences (within days) +++ ++ +++ ++ −
Capacity to link sequential experiences (across days) − + ++ − +++

Note. A minus suggests that the method is relatively ineffective at achieving the conceptual goal, whereas pluses
indicate relative effectiveness (which increases as pluses increase) at achieving the conceptual goal. CRA = continuous
rating assessments; DRM = day reconstruction method; ESM = experience sampling methodology.
Copyright American Psychological Association. Not for further distribution.

within days versus a goal to link experiences across experiences—the least effective approach is continuous
days based on the current state of the intraindividual rating assessment. This should come as no surprise:
literature. That is, many studies now seem focused The purpose of CRA is to take an in-depth look into
on linking immediate, within-day experiences a particular event in isolation, rather than obtaining
together (e.g., Beal et al., 2013; Gabriel et al., 2014; a representative sample of experiences across several
Rosen et al., 2016; Trougakos, Beal, Cheng, Hideg, days. In contrast, the other varieties of ESM and diary
& Zweig, 2015), whereas others focus on linking designs are capable of accomplishing the goal of
aggregate daily experiences or experiences lagged representativeness to varying degrees. This also should
across the natural break of each day (e.g., Beal & come as no surprise, as ESM methods were derived in
Ghandour, 2011; Foo, Uy, & Baron, 2009; Scott & large part as a way of introducing ecological validity
Barnes, 2011; Wang et al., 2013). into an area of social research that often relied on
experiments that were unlikely to capture experiences
Representativeness.   When we consider representa­ as they were lived (Csikszentmihalyi & Larson, 1987;
tiveness in intraindividual research, we are focusing cf. Aronson & Carlsmith, 1968). Interestingly, it is
on the extent to which researchers are capturing difficult to determine the extent to which the DRM
the lived experiences of individuals in a variety of accomplishes this goal. On one hand, it samples a
contexts (e.g., Weiss & Rupp, 2011). This is often single day from each participant, leaving glaring
why researchers ensure that their survey period spans holes in a variety of experiences each of us have
several days and weeks, with surveys administered over even a few days’ time. On the other hand,
at different times: It is assumed that individuals’ samples for DRM studies are typically large, so they
experiences change from day to day and from event may achieve a representative sample of experiences
to event, and researchers are aiming to capture an across people rather than within. The downside,
accurate reflection of what happens to individuals however, is that despite a potentially large range of
across time. As such, if researchers were to only survey experiences across people, individual antecedents
at a particular time of day (e.g., the end of the day), and consequences of those experiences will be
or on a particular day of the week (e.g., Monday or limited to what occurred on a single day.
Friday), they would run the risk of missing out on Of the ESM permutations, there are differences
both typical events (e.g., what happens to most in their capacity to gain a representative sample of
people on a typical day), and, on the other end of the experiences. Daily ESM studies, which emphasize
continuum, extreme or atypical events. mental aggregations of states experienced during
If we examine the first row of Table 13.1—the goal the day (e.g., “To what extent have you been
of capturing a representative sample of an individual’s satisfied with your job today?”), as well as episodic

311
Beal and Gabriel

ESM studies (e.g., “To what extent have you been method features the reporting of immediate states
satisfied with your job since the last survey?”), are randomly sampled throughout the day. As such, it
somewhat limited in providing a representative is unlikely to capture a biased set of experiences
depiction, as the variability of experiences that that would occur only if a certain type of event were
occurs within a day or episode is funneled into sampled or if surveys were only administered at
a single score. As a result, these methods cannot a particular time. Yet, it is interesting to note that
capitalize on within-day or within-episode the approach to ensure a representative range of
variance to obtain a full picture of one’s life as experiences has not been discussed in great detail.
it is lived. That said, these methods have their Most ESM discussion (e.g., Beal & Weiss, 2003;
merits. First, it is unlikely that unusual states Bolger, Davis, & Rafaeli, 2003; Csikszentmihalyi
within a day or episode would not be reflected & Larson, 1987) provides descriptive accounts of
in some aggregation of those states. For example, what researchers typically do when conducting
suppose that an employee experiences typical ESM, rather than suggesting a theoretically driven
Copyright American Psychological Association. Not for further distribution.

levels of job satisfaction throughout most of the rationale for study design choices.
day but in the afternoon receives news suggesting By identifying the key elements of ESM, Beal
that a promotion is likely. The influence of this (2015) provided a foundation from which such a
pivotal event is likely to be borne out in the day’s theoretically driven approach might be established.
assessment of job satisfaction. Capturing this peak With respect to representative sampling, the goal
level of job satisfaction helps the entire sample of in determining the sampling strategy, in terms of
daily ratings become more representative of the full signals per day and total duration of the study, should
range of job satisfaction that one might experience lie in an assessment of how the experience under
(cf. Fredrickson, 2000). The same process would examination changes over time. If the experience
be true for episodic ESM, but perhaps to an even (e.g., states of happiness) fluctuates greatly over
greater extent, as aggregations represent only single brief periods of time, then a brief period of time
episodes within a day rather than the entire day. We could produce a representative sample of an
should note, however, that the process governing individual’s experiences. The focus on variability of
which states are reflected in an individual’s episodic the experience in question and the capacity to obtain
or daily report is a question for which answers are representative samples from potentially small sets
only at an early stage (e.g., Miron-Shatz, 2009). of experiences has been emphasized most ardently
As a result, elements of experience that color one’s in the qualitative methods literature (e.g., Glaser &
daily or episodic reports may help increase their Strauss, 1967; Gobo, 2004), and we suspect that this
representativeness, but these elements may also type of discussion is clearly applicable to ESM as well.
decrease it. An additional question that future As an example, consider a state such as boredom.
research must address is the predictive validity of Although there may be a number of different forms
different levels of aggregate responses: There may of boredom, it seems fair to suggest that, relative
be differences between immediate satisfaction
to states of anger or happiness, boredom may not
aggregated to the day and one rating of daily
exhibit a tremendous degree of variability over time
satisfaction, but if both approaches are related like
within a particular individual. As a result, obtaining a
outcomes of interest are, then these details may
representative sample of boredom states would likely
not matter. In contrast, if these approaches are
require far fewer instances of boredom than would be
differentially predictive, then understanding the
necessary to obtain an equally representative sample
respective patterns of prediction may be pivotal.
of anger or happiness states.
That said, the method that best accomplishes
the goal of representativeness is the original Dynamics.   As displayed in second row of Table 13.1,
momentary ESM described by Czikszentmihalyi researchers also may have a goal of determining what,
and colleagues (e.g., Csikszentmihalyi & Graef, if any, pattern of change exists in a given variable. In
1980; Csikszentmihalyi et al., 1977), given that this longitudinal panel designs, where change is examined

312
Looking Within

over a longer period of time but with relatively few 2011; Benedetti et al., 2015) and other relevant
periods of observation, latent growth models often constructs (e.g., Rosen et al., 2016).
provide an estimate of the overall pattern of change. The final method that could be used to capture
In intensive repeated measures designs such as ESM dynamic patterns of a conceptual variable is
or CRA, patterns of change tend to be much less stable moment­ary ESM. This approach has both advantages
and far more complex. Beal and Weiss (2003) and and disadvantages. For example, it is the only
Beal (2015) discussed how anticipated patterns might approach capturing immediate states that has the
be examined using trends, cycles, and autoregressive capacity to examine longer term dynamic patterns
effects, while Beal and Ghandour (2011) provided a (i.e., across days). CRA captures dynamics of
substantive example using daily positive and negative immediate states as well, but necessarily does so
affect. Perhaps the most important factor to consider over a very brief period of time. The issue that arises
when evaluating the extent to which a given method for momentary ESM, however, is that the random
can accurately and comprehensively map the dynamic dispersion of observations throughout the day
Copyright American Psychological Association. Not for further distribution.

patterns of a conceptual variable is the level of makes it difficult to understand the continuity of a
aggregation for the variable. particular dynamic state. Continuing with affective
Take, as an example, affective states. When state as the example, a momentary ESM study might
construed as an immediate state, it is likely to capture a high level of anxiety at 10:03 a.m., and a
fluctuate a great deal over potentially very brief similarly high level of anxiety reported at the next
periods of time. Gabriel and Diefendorff (2015) signal occurring at 12:53 p.m. This second high
demonstrated this very effectively using CRA to map point for anxiety could reflect the continuation of
affective states during customer interactions lasting the previous state, or it could represent the peak
less than 5 minutes. Indeed, at the most immediate of an entirely new state that emerged during the
level of aggregation, CRA is by far the best equipped intervening survey period. Because momentary ESM
to examine dynamic patterns of volatile variables, cannot assess the true variability of the continuous
and it can be used to help researchers understand— stream of immediate affective experience, it may
perhaps in an exploratory manner—the rate at which not provide a completely accurate description of
a construct of interest changes within an event. Yet, its dynamic patterns. Ironically, neither episodic
affective state is a good example of a variable that approaches, DRM, nor daily approaches suffer from
has been successfully conceptualized at numerous this dilemma, as they are focused on more aggregate
levels of aggregation. For example, key events are forms of affective experience. That is, episodic
known to give rise to episodically experienced ESM, DRM, and daily ESM provide reasonably
emotion (Russell & Feldman Barrett, 1999; Weiss & good depictions of the dynamics of episodically
Cropanzano, 1996), and describing the key features of experienced states and daily states. However,
such episodes allows for a meaningful and predictive they would provide fairly poor depictions of the
level of aggregation (e.g., Beal et al., 2006). To dynamics of immediately experienced states, but of
course that is not the goal of these methods.
the extent that emotion episodes are experienced
in a patterned manner over time, episodic ESM, Linking within days.   Perhaps one of the most
DRM, or combinations are the most appropriate common goals of researchers using some form of
way in which to capture such a pattern. Similarly, within-person design is to connect one conceptual
numerous papers have relied on daily aggregation of variable with another one within the same day.
affective state (i.e., asking people to report on their Too often, this goal is implemented using a con­
affective experience for the day) and have shown that current within-person correlational analysis
this level of aggregation also reliably predicts other strategy. Although there are conditions under which
constructs similarly assessed at the daily level (e.g., concurrent relations might be feasible or preferable,
Judge, Scott, & Ilies, 2006; Sonnentag & Zijlstra, if causal inferences are sought, this approach is
2006). This approach also has demonstrated reliable certainly suboptimal (Edwards, 2008). That is,
dynamic patterns of affect (e.g., Beal & Ghandour, even if multiple assessments of constructs are

313
Beal and Gabriel

taken across several days, if the researchers analyze also connect constructs within a single episode.
relationships assessed at the same point in time, the For instance, Gabriel and Diefendorff (2015)
same issues associated with common method bias used a series of lagged variables to examine the
and correctly specified causal order still plague DRM causal relationships among felt emotions, emotion
and ESM assessments. Instead, researchers should regulation, and ratings of emotional performance
be pursuing these goals using lagged effects with (i.e., vocal tone). As such, even the most micro
appropriate temporal controls (Beal, 2015; Beal & of intraindividual approaches can be used to link
Weiss, 2003). Although this sounds easy enough, constructs together across multiple time points.
the details of this approach need to be carefully As these examples demonstrate, both CRA
considered when an ESM study is being designed. In and episodic ESM are effective at capturing the
particular, one must consider the pacing of changes connections between a day’s events, experiences,
not only in both conceptual variables separately but and behaviors. Clearly, daily ESM cannot serve
also in how they fluctuate together. For example, this role, as it necessarily aggregates within-day
Copyright American Psychological Association. Not for further distribution.

if someone is interested in how task autonomy experiences. Momentary ESM does have the capacity
predicts feelings of psychological vitality, one must to link experiences within days, but as this method’s
consider not only the dynamic properties of both momentary emphasis cannot provide a depiction of
variables separately (perhaps autonomy fluctuates the continuous stream of experience within a given
on a daily basis, while vitality fluctuates on more day, there is again the possibility that momentary
of a momentary basis) but also what measurement states will be more difficult to connect with prior
approach would be best to capture their covariation. or subsequent momentary states. As episodically
When possible, it is usually advantageous to measure framed items are designed to capture the essence
all variables at the fastest pace—there might be of what was experienced (at least from a subjective
less within-person variation in the less dynamic point of view), this approach likely represents a
construct, but it is likely a worthwhile endeavor to better option for the purpose of linking experiences
demonstrate this characteristic empirically. within days. Although the DRM also holds this
Interestingly, researchers are beginning to advantage, it has an additional disadvantage.
pay more attention to linking constructs within Specifically, as reports of within-day experiences
days in ESM research. For example, Rosen et al. are solicited at the same time, this method seems
(2016) studied the relation between incivility and particularly likely to be subject to consistency and
depletion. Specifically, using an episodic approach, other forms of mono method bias (see Podsakoff,
the researchers assessed exposure to incivility MacKenzie, Lee, & Podsakoff, 2003).
in the morning (i.e., Time 1) and related it to a Linking across days.   Beyond assessing constructs
concurrent assessment of depletion (i.e., Time 1). within a day, researchers are further examining
Rosen and colleagues then sought to link depletion how constructs captured via ESM may spill over
to subsequent acts of incivility toward others later across days. For instance, it may be that a construct
in the workday (i.e., Time 2). By doing so, the captured in the evening spills over to predict a
authors were able to examine more downstream work-related attitude the next workday. When this is
consequences of incivility and subsequent depletion being considered, it is likely the case that episodic or
that occurred within the same workday. Likewise, daily ESM approaches would best lend themselves
Koopman, Lanaj, and Scott (2016) assessed how to these questions, as researchers need to carefully
engaging in organizational citizenship behaviors consider when certain experiences may happen;
in the morning predicted increased positive affect using a momentary approach could miss key events
and decreased work-goal progress in the afternoon, (e.g., a negative or positive interaction at home after
allowing them to delve into the “bright” and “dark” work; Butts et al., 2015). CRA and DRM—given
sides associated with engaging in citizenship their focus on single events and single days—are
behaviors throughout the workday. As a final unlikely to fit questions tied to linking constructs
example, using a CRA approach researchers can across days. Of course, alterations of DRM that

314
Looking Within

extend across multiple days (e.g., Bakker et al., 2013) operates across a broad and representative range
would be less subject to these issues and would of contexts, particularly contexts that are temporal
potentially be effective for this goal. in nature. They are geared more toward filling in
Areas of research that are seeing an increase in particular, but perhaps idiosyncratic, sections of a
designs that link experiences across days include much larger canvas—each effort provides a piece
research at the work–family interface. For example, of the picture, but little sense of the whole can be
in the study previously described by Butts et al. made until a great many pieces have been added.
(2015), the researchers not only assessed experiences This approach is certainly germane to within-
with after-hours e-mail as they happened but also person organizational research as much as it is to
linked these evaluations to important outcomes more macro areas of research, but we feel that the
the next workday, including perceptions of work– suite of methods described here have the potential
family conflict. In research related to workplace to be linked in a programmatic way that builds
recovery, Sonnentag, Binnewies, and Mojza (2008) understanding in a more coherent fashion.
Copyright American Psychological Association. Not for further distribution.

had participants evaluate different types of recovery To accomplish this goal, we will be using an
experiences (e.g., psychological detachment, example of an aspiring social science researcher
relaxation, and mastery experiences) they had in the who would like to understand more about the
evening after work and linked these experiences to immediate experiences of positive self-conscious
sleep quality and affect the next morning. emotion and its links to how discriminatory
behaviors are interpreted by low-status individuals,
Linking Information Across or literally: pride and prejudice. Although one
Within-Person Methods approach would be to examine trait levels of pride
Although each intraindividual design associated and global perceptions of discrimination, our
with ESM and/or CRA can be viewed as either/or, researcher believes that the manner in which these
we argue—as do Gabriel et al. (2017)—that the two constructs connect on an immediate, dynamic,
methods should be viewed in a more integrated and within-person basis is likely to yield greater
manner to help researchers not only to better insights. As neither construct is well understood
understand how relationships change depending on through a within-person lens, an initial exploratory
the level of analysis but also to better understand investigation seems warranted.
complex phenomena of interest. Gabriel et al.
Step 1: Exploration of temporal properties.  Given
(2017) suggested that researchers may be able to,
that most scientific efforts in the psychological and
for instance, start with ESM research to identify the
organizational sciences are based at the levels of the
prevalence of a type of event and then use CRA to
individual, the group, or the organization (Kozlowski
better unpack the dynamics that may or may not
& Klein, 2000), it is at least understandable, if not
occur within that event. However, as we describe
predictable, that the dynamic properties of within-
below, it is likely that researchers can move back and
person phenomena will not be easily specified at the
forth across the various methods.
earliest phases of study design. For organizational
Proposal for an integrative, programmatic approach. researchers in particular, tackling a within-person
In our collective experience, it is clear that many topic will often require some initial exploration.
research efforts may not follow a programmatic, Although it might seem reasonable to start this
phenomenon-based developmental process. Instead, exploration at the lowest level of aggregation
we often conduct one or more studies in an attempt possible (e.g., using CRA), we argue instead that
to answer a specific practical question, to clarify a starting at higher levels of aggregation is likely
particular theoretical ambiguity, or to combine two more useful. As we just stated, this field is not
or more theoretical accounts of similar effects or intimately familiar with intraindividual processes and
sets of effects. There is nothing wrong with these dynamics, and so mapping out the most immediate
approaches, yet they do not often lend themselves and momentary instances of a given construct may
to a detailed account of how a particular process not be wise. More importantly, a more aggregated

315
Beal and Gabriel

approach can provide information relevant to the discrimination is quite common for minorities and
baseline frequency of a particular experience, state, stigmatized groups; Leslie, King, Bradley, & Hebl,
or behavior, with the modest assumption that 2008), so starting at a broader level and attempting
the occurrence of most focal constructs will be to summarize the frequency of occasions is advisable.
memorable enough within the day on which they A number of other organizational scholars have
occur. suggested that—at least for the field of organizational
As an example, consider the construct of pride research and likely beyond—there is a need to
being examined by our trusty novice to within- include this initial attempt at uncovering dynamics
person approaches. Pride, as it is based in specific (George & Jones, 2000; Mitchell & James, 2001)
contexts of socially valued events characterized before meaningful and temporally informed theory
by personal responsibility (Tracy, Shariff, & can be applied. We echo this, but we recognize that
Cheng, 2010), may not have a noticeable regular running a study of this complexity for no purpose
occurrence for many people (perhaps excepting other than documenting temporal features of the
Copyright American Psychological Association. Not for further distribution.

hubristic individuals; see Tracy & Robins, 2007). phenomena being examined is unlikely to be met
Of course, given the human tendencies toward with open arms from researchers who have too
self-enhancement and self-serving attributions, we many pragmatic concerns and resource limitations to
might also reasonably expect at least modest levels invest such effort. Indeed, one of the reasons why
of pride on a regular basis. Not knowing what to the combined DRM/daily ESM approach is useful
expect, it might be advisable for our researcher to is that it allows researchers to go beyond simple
begin her efforts with the version of the DRM that documentation of temporal properties to examine
blends with daily ESM. As the DRM is typically substantive questions concerning how two or
more intensive than other ESM surveys, the more dynamic variables are linked together.
duration of the study might be reasonably limited We are aware, however, that this sort of initial
to a week. This approach carries with it several study is unlikely to provide definitive evidence either
advantages to an initial mapping of the dynamic and for highly accurate within-day temporal properties or
substantive properties of pride experiences. First, as for evidence of causal order. With respect to accurate
an emotional experience, it is likely to have episodic within-day temporal properties, this approach relies
properties that would be captured better by the DRM on the DRM and the recall of the day’s episodes along
than daily ESM alone. At the same time, if pride with their qualities and durations. Diener and Tay
experiences carried forward into the next day (or (2014) pointed out that when the DRM approach is
longer), including a daily ESM component would compared with momentary ESM approaches, the
facilitate detection of such lagged effects. Another DRM is estimated to be only modestly related to
advantage of combining approaches is that an initial temporally matched scores on ESM, and they go
mapping of dynamic patterns (even within-day) further to suggest that “the size of these estimates
will lead investigators to use ESM where this is
would be feasible using the time-sequenced aspects
feasible” (p. 258). Furthermore, as we have noted
of the DRM (Kahneman et al., 2004). Although
earlier, attempts to link two or more variables within
Table 13.1 suggests that neither of these methods are
days using the DRM is likely to contain substantial
particularly well suited for the goal of obtaining a
amounts of method variance. As people are asked
representative sampling of experiences, as we noted
to recount the episodes of their day and then rate,
earlier, the combination of the two methods might
for example, feelings of pride as well as experiences
offset their respective shortcomings here and be an
of prejudice, then any implicit narrative, transient
effective way to address this goal.
mood state, or desire to be consistent in their
This approach might also be useful for exploring
responses will serve to obscure the true connection
the dynamic properties of prejudice experienced by
between these variables.
low-status individuals. Again, it is unclear how often
instances of prejudice might be perceived (though Step 2: Informed application of within-person
research suggests that both perceived and objective design.   As these issues suggest, researchers should

316
Looking Within

not stop with an initial combined DRM/daily ESM same type of jobs from the same organization), then
study. Instead, they should treat the information it might be possible to identify common episodes
gleaned as suggestive of the temporal properties and that form the basis of the within-day sampling
potential links between two or more within-person regimen (e.g., mealtime rushes for wait staff; Beal
constructs of interest. With this information in et al., 2013). If there is greater heterogeneity of
hand, researchers should be better able to select the daily activities, then one might resort to more
most appropriate within-person design for the next typical daily temporal structures. For example,
study. As Step 1 provides only information aggregated the first survey might be completed midmorning
to the episodic level, then it seems unlikely that a CRA and refer to activities involved in waking up,
study would follow. As we suggest below, however, preparing for the day, and arrival at school or work.
incorporating this method into a programmatic effort The second survey could be completed prior to
is likely essential to a full understanding of many lunch and cover activities since the midmorning
within-person phenomena, and it is the centerpiece survey. A third survey might occur midafternoon,
Copyright American Psychological Association. Not for further distribution.

of the third step described below. At this point, covering the postlunch period of time, and a final
however, if the phenomena under investigation survey could be completed at the end of the day,
revealed substantial or likely variation at the summarizing experiences since the midafternoon
within-day level, then either the episodic ESM survey. Although these surveys are not necessarily
or momentary ESM approaches would be likely aligned with the subjectively experienced episodes
candidates for a second study. of each individual, targeting a sample’s collective
A consideration of the advantages of these activity structure can provide a reasonable degree of
two designs taken from Table 13.1 reveals under overlap with episodic structures likely experienced
what circumstances one might select episodic by many individuals. Still, another option would
or momentary ESM. Although both methods are be to use an event-contingent approach focused
somewhat effective for each research goal, a concise on particular episodes of interest, such as the
summary suggests that if obtaining a representative experience of pride or perceptions of prejudicial
sampling of experiences is paramount, the clear behavior from others. Creative combinations
choice would be a momentary ESM design. For all are also possible, such as the blending of signal-
other research goals described in the table, however, contingent and event-contingent approaches (e.g.,
episodic ESM holds a slight advantage. The primary participants are instructed to complete a survey
reason for this advantage is the capacity of episodic each time they perceive prejudicial treatment while
ESM to capture the continuity of an experience both also completing scheduled or signaled surveys to
within and across days. Momentary ESM, with its capture episodic experiences of pride).
emphasis on snapshots of immediate states, has the Regardless of the approach taken, items would
potential to mischaracterize this continuity. ask participants to summarize over the intervening
For example, let’s say that Step 1 revealed to our episode or period of time, allowing lagged or cross-
novice researcher that pride experiences ranged lagged effects to be more readily detected using
from frequencies of multiple times per day to once appropriate modeling choices (Beal, 2015). This
every few days, whereas experiences of prejudicial approach could provide reasonable answers, for
treatment from higher status individuals occurred example, to questions such as whether the prior
multiple times per day for all of the low-status experience of pride buffers subsequent perceptions
participants. Despite the slightly slower pacing of of prejudicial behavior, or whether prior perceptions
pride experiences, an episodic ESM approach is well of prejudicial behavior limit the extent to which
matched to both variables. A well-designed episodic pride is subsequently experienced. This capacity to
ESM study would carefully consider the possible link separate episodes of two or more variables while
or likely episodic structure of the sample’s day. If controlling or examining dynamic patterns of the
the sample is relatively homogenous with respect modeled variables is a key strength of the episodic
to prominent daily activities (e.g., they all hold the ESM approach.

317
Beal and Gabriel

One possibility that we have yet to consider in part of the ESM survey (e.g., framing of items
Step 2 is what to do if the initial step was suggestive or inclusion of manipulated instructions). Together,
of a slower within-person process that occurred the inclusion of a time frame well matched to
more on a day-to-day basis than on a within-day the variables of interest along with the inclusion
basis. In many ways, this discovery would place the of one or more manipulated independent variables
researchers at an advantage, as Step 1 happened to would generate a particularly potent and rigorous
be well suited to the dynamics that were identified. study.
Indeed, one might reasonably ask whether our
Step 3: Isolating and dissecting crucial episodes.  
researcher should stop with Step 1. Although doing
The episodic approach outlined above, particularly
so certainly seems ideal for minimizing an effort-
when it is preceded by an initial temporal
to-reward ratio, we suspect that the concerns with
exploration of the relevant variables, provides a
this approach noted earlier limit the theoretical
robust examination of within-person dynamics.
contribution from Step 1 to the point that it would
Copyright American Psychological Association. Not for further distribution.

It is often difficult to delve more deeply within an


be difficult to reach the bar currently set by many
episode, however, particularly when intervening
journals. In which case, what should be the next
mechanisms (i.e., mediators) are likely important
step? Our answer involves adding at least two
but occur quickly either after the experience of
components of increased precision and rigor to the
the predictor variable or immediately prior to the
initial daily ESM design. One component would criterion variable. Such mechanisms are quite
be to provide a better match to the discovered time common, as cognitive or affective reactions often
course. Specifically, the initial combined DRM/ follow an event of interest very rapidly (e.g., Sander,
daily ESM design is likely to have extended only Grandjean, & Scherer, 2005). Behavioral responses
for a few days. As we noted above, the DRM places to these reactions also are often close to immediate.
an increased burden on participants and typically Thus, a comprehensive mapping of an entire event–
demands a larger sample size. Trade-offs in study interpretation–action process might transpire over
design therefore suggest that such a design would the course of seconds or minutes.
be difficult to sustain over long periods of time. If, Clearly, none of the ESM approaches (or DRM)
however, the initial study suggests a longer time are capable of describing processes of this sort, yet
frame for the dynamics of the variables under our lives are replete with experiences just like this.
investigation, then using just a daily ESM design Indeed, scholars have begun to try and tease apart
would enable researchers to extend the duration events, their immediate reactions (e.g., depletion),
of the study, allowing for a more comprehensive and subsequent behaviors (e.g., Rosen et al., 2016),
examination of temporal patterns. but these designs often involve an assumption
The second component would be to include of which part of the process is occurring faster
an experimental manipulation that would or slower than the other (i.e., assuming that a
strengthen causal inferences implied in Step 1. particular timing of an ESM survey is appropriate).
Although experimental manipulations are rare in For instance, in the Gabriel and Diefendorff (2015)
ESM research, a recent review of this literature study referred to previously, participants were
recommended their implementation (Beal, 2015). put through a call center simulation in which
In many ways, this approach reflects an ideal they were exposed to a hostile customer; this
combination of field and lab studies: ESM provides context was chosen to build on past emotional
the representative, naturalistic context of in situ labor research that had focused on incivility in an
experiences, and the inclusion of within- or between- experimental context (e.g., Rupp & Spencer, 2006),
person manipulations provide an increased level of but could not properly test the dynamics of the
rigor absent from field work. Such manipulations event. By assessing the felt emotions in response
could be implemented outside of the ESM survey to incivility, the regulatory response participants
(e.g., employees begin ESM either in a pretest– had to the incivility (i.e., how much participants
posttest training group or in a control group) or as a regulated their emotions), and their performance

318
Looking Within

(i.e., third-party ratings of vocal tone), mapping the MOVING TO THE BETWEEN-PERSON LEVEL
event–interpretation–action process was possible.
Although the aforementioned discussion focuses
Indeed, using a cross-lagged panel design, Gabriel
largely on how to move across several sublevels
and Diefendorff were able to temporally separate
within the microlevel of analysis, there is room
and identify which processes came first (e.g., felt
to consider how our understanding of temporal
emotions always → emotion regulation, but not vice
dynamics may coalesce and shape person-level
versa). In their review of CRA, Gabriel et al. (2017)
attitudes and behaviors. An exhaustive discussion
further provide several examples and applications
of every possibility is beyond the scope of this
of CRA, from examining the unfolding dynamics of
chapter, but we detail some ways that researchers
performance feedback exchanges, to understanding can integrate within- and between-person research
how job applicants experience recruitment to create a more informative story.
information as they are being exposed to it. Again,
all of these instances involve a deep dive into one Composition and Compilation Models
Copyright American Psychological Association. Not for further distribution.

singular, salient episode for employees. As was true with early efforts to specify multilevel
In the case of assessing experiences associated structure and function (Kozlowski & Klein, 2000),
with pride and prejudice, CRA could help glean particular care should be taken to consider how
several insights not fully captured with the various within-person states, perceptions, and behaviors
ESM or DRM approaches. For example, when the emerge at higher levels of analysis (e.g., homologous
possibility that pride might influence perceptions versus differential effects at higher levels of analysis).
of prejudice is examined, participants could be Unfortunately, the within-person literature has paid
exposed to a manipulation of pride (vs. neutral or fairly little attention to such issues (see Gabriel et al.,
other affective state) and then placed in a carefully 2014, as an exception). For instance, concepts such
crafted simulation in which another participant as composition and compilation—commonly used
(confederate) engages in subtle (or perhaps not- in the multilevel literature focusing on groups and
so-subtle) forms of discrimination. Participants teams—are scarcely mentioned in the within-person
would provide continuous ratings on one or more literature. In part this is likely due to the lack of
dimensions of their experience (e.g., perceptions connection between these two multilevel areas of
of fair treatment, current affective state, current research (i.e., terms developed in one area may not
levels of task, or other forms of motivation). easily transfer to another), but the larger issue is that
Importantly, this analysis provides information focused consideration of what these concepts mean
about multiple specific behaviors or events that is rare in the within-person literature. To be clear,
occur within a single episode that would be and from a within-person perspective, composition
difficult to obtain from the aggregate reports of models would suggest that momentary or episodic
episodic ESM. Gabriel et al. (2017) referred to states, perceptions, and behaviors function in a
these shorter instances as microevents within a similar way (i.e., homologously, isomorphically)
larger event, meaning that even within one single when they emerge at the person level of analysis. As
episode, different microevents may unfold that give a result, individual-level constructs comprise one
way to important temporal dynamics. Likewise, of several potential aggregations of states measured
scholars in social psychology have used CRA to at the immediate, episodic, or daily level; that is,
assess how specific emotions—such as anger or a composition model specifying the emergence of
happiness—occur when participants are exposed conscientiousness at the individual level might
to stimuli such as images or short videos (e.g., consist of the average of conscientious behaviors
Mauss et al., 2005). Again, the goal in Step 3 is to measured over repeated episodes for a given
dissect broader episodes into microevents that can individual over several days (cf. Fleeson, 2001).
aid understanding in both (a) why constructs vary Chan (1998) described a number of possible
momentarily and (b) how processes interrelate composition models for group-level phenomena
within a given event. that emerge from properties of individuals, and

319
Beal and Gabriel

Kozlowski and Klein (2000) clarified and expanded To the extent that deviations from the prototypical
upon many of these more specific forms. These states, perceptions, and behaviors are of interest
models apply equally well to the within-person as phenomena that emerge at the individual level,
domain, though the implications of their tenets Kozlowski and Klein (2000) would argue that
differ in interesting ways. For example, an additive such complex aggregations are better considered
model would suggest that the higher level (i.e., compilation models, as they capture something
the individual level) is a simple sum or average distinct from the immediate experience itself.
of the lower level parts (i.e., relatively immediate For example, dispersion models shift the emphasis
states, perceptions, and behaviors), regardless of the away from stable averages of life’s experiences
variance of those parts. In contrast, a direct consensus and toward the amount of variation that occurs
model requires evidence of relatively homogeneous around one’s average experience (something
lower level parts as justification for aggregation. unlikely to be imparted by a self-report of one
Thus, someone who varied greatly in terms of particular experience). Although within-person
Copyright American Psychological Association. Not for further distribution.

conscientious behavior from day to day or event to research examining this sort of compositional
event could not be described by his or her average model is scarce, there are examples. For instance,
level of conscientiousness under a direct consensus Beal and his colleagues (Beal & Ghandour, 2011;
composition model, given that evidence of stability, Beal et al., 2013) have examined the concept of
perhaps using ICCs or agreement indices, would affect spin, which is a way of capturing variability
need to be provided first, and these statistics would in affective state. Likewise, Scott and colleagues
likely indicate that researchers should not engage in have begun to consider within-person variation as a
aggregation. However, such an individual could be meaningful construct at the between-person level of
described by an average level of conscientiousness if analysis, examining variables such as emotional labor
an additive model were put forth. variability (e.g., Scott, Barnes, & Wagner, 2012) and
Chan’s (1998) referent-shift model has an daily variability in justice (Matta, Scott, Colquitt,
interesting translation as a composition model for Koopman, & Passantino, 2017).
within-person properties. In this case, it implies Finally, process-oriented compilation models
that individuals repeatedly describe a particular emphasize complex connections among parts of
emergent property of themselves over time. That the whole that produce emergent properties. For
is, rather than asking a number of people whether within-person phenomena, this model would refer to
their group acts in a conscientious manner, it connections among states, perceptions, and behaviors
might instead entail asking an individual about his over time. For example, Glomb, Beal, Yang, and
or her level of conscientiousness repeatedly over Bhave (2014) examined negative emotional inertia
time. One implication of this model for within- as a factor in understanding how workers react to
person phenomena is that it will clearly shift the negative events. In this case, negative emotional
bulk of the variance away from the moment-to- inertia was specified as the autoregression of
moment variations of the target characteristic negative affect (i.e., on immediately prior negative
and toward the more enduring and stable features affect). As such, it represents a complex connection
of the evaluated self (Kahneman & Riis, 2005; between two consecutively experienced states. Such
Robinson & Clore, 2002). The interesting question models have been discussed in more detail by Wang,
that arises from considering this model is whether Hamaker, and Bergeman (2012) and in the emotions
it is the recalled, stable features of identity that literature, in particular (e.g., Kuppens, Oravecz, &
best characterize an emergent individual-level Tuerlinckx, 2010).
property, or whether it is best characterized not
only by what is typical but also by the vagaries, Emergent Properties as Perceptual
peaks, and rock bottom of lived experience Versus Mathematical Aggregations
(e.g., “peak and end” information; Fredrickson Thus far, we have discussed details and considerations
& Kahneman, 1993). in situations when researchers might wish to aggregate

320
Looking Within

within-person phenomena to the between-person conscientious behavior. The interpretation at this


level. However, we have not considered whether such level of analysis would therefore reflect how changes
an effort would be either reasonable or advisable. As in an individual’s conscientious behavior relative
an example, let’s again consider the personality trait to his or her typical level of conscientiousness (i.e.,
of conscientiousness. At the within-person level, conscientious behavior across all days of the study)
there is now ample evidence that within-person predict an outcome variable of interest (e.g., ratings
fluctuations in conscientiousness are substantial in of performance). To examine the effects at a higher
terms of both frequency and magnitude (Fleeson, level of analysis would therefore require adding the
2001, 2004). As a result, we can meaningfully ask means (or other aggregations of conscientiousness)
about the antecedents and consequences of more into the model as a person-level predictor variable.
immediate conscientious behavior. Moreover, Furthermore, given the likely statistical dependence
we now have a second viable means to assess of means with other aggregations (e.g., variance
conscientiousness as a trait. The first method, of and skewness), the means would likely need to
Copyright American Psychological Association. Not for further distribution.

course, would be through a traditional personality be incorporated alongside other aggregates at the
questionnaire; the second method, however, would between-person level of analysis. For example,
be one of several potential aggregations of the examining whether variability in conscientious­
within-person conscientiousness states along the ness predicted average levels of performance for
lines of the composition and compilation models individuals in a sample would require first controlling
outlined above. Given the discussion earlier in for average levels of conscientiousness. The reason for
the chapter about the meaning of within-person this step is that, particularly when using scales such
phenomena captured immediately, episodically, as the ubiquitous Likert-type scale, higher or lower
or daily, it should by now be clear that these two average scores are likely to reduce variance due to
methods could produce fairly distinct estimates the upper and lower boundaries of the scale (Beal
of personality, with the first reflecting the use & Dawson, 2007). Thus, without first considering
of semantic memory in constructing stable the influence of average conscientiousness, one
self-perceptions or self-evaluations, and the might interpret conscientiousness variability
second reflecting a prespecified mathematical as having predictive validity, even though this
aggregation of multiple, temporally dispersed validity is potentially due to the average level of
self-perceptions of personality derived primarily conscientiousness.
from episodic memories obtained close to the In addition, the aggregation of repeatedly
experienced conscientious behavior. Research has measured states often carries with it differences in
supported the notion that these two estimates of the psychometric properties of the measures used,
personality are strongly correlated but empirically particularly as they compare to the best examples of
distinct constructs (Fleeson & Gallagher, 2009; measures used for assessing traits (e.g., measures of
uncorrected meta-analytic rs between .42 and .56
the Big Five or other such perceptual aggregations
[.48 for conscientiousness specifically]). Moreover,
of individual differences). Specifically, because
other aggregations, such as variability in conscientious
ESM or CRA methods are assessed repeatedly, it
behavior, provide even more distinct descriptions of
is difficult to require participants to respond to
personality.
lengthy, broad-domain measures without creating
Statistical considerations.   Although the conceptual undue burden—and consequent problems with data
connections of these elements of personality to quality. This dilemma seems to have been handled
other constructs seem like exciting new ground in ESM studies in several ways, but broadly there
to explore, there are a number of more pragmatic are two approaches: Assess a narrower domain of
considerations. For example, most multilevel the construct of interest, or use only a few items to
models examining within-person states use cluster- assess a broad domain. The first approach allows
mean centering (i.e., within-person centering) researchers to use scales that will often achieve
for within-person predictor variables such as acceptable levels of internal consistency reliability,

321
Beal and Gabriel

but the cost is that the full domain of the construct reflect what Kahneman and Riis (2005) referred to
cannot be captured. In these cases the onus is as the “remembering self,” episodes or personality
on the researchers to provide evidence that the states assessed repeatedly across a representative
narrower domain assessed is not overstated in period of time and aggregated to the level of the
its relevance to the full domain of the trait-level individual reflect what Kahneman and Riis referred
construct. Consequently, ESM studies taking this to as the “experiencing self.” For instance, in the
approach will have a more targeted focus in their world of work and organizations, the differences
research goals and conclusions. between these two reflections might be fairly
The implications of the second approach depend, important. Methodologically, it seems plausible
at least in part, on the viability of assessing broad that indices of the experienced and evaluated selves
domains with a small number of items. In some vary in terms of the common method biases that
cases, broad overall assessments can still achieve might be present. For example, linking indices
acceptable levels of reliability with only a few of the experienced self to other, more evaluative
Copyright American Psychological Association. Not for further distribution.

items (e.g., job satisfaction; Ilies & Judge, 2002; constructs (e.g., job satisfaction) might be less likely
Judge & Ilies, 2004). In other cases—particularly to be influenced by consistency motifs or implicit
when a higher order construct composed of theories that might otherwise inflate these effects
multiple subdomains is being assessed—the result (cf. Podsakoff et al., 2003). In contrast, indices
is relatively poor reliability estimates for the brief based on experiences might be more likely to be
scales that are used (e.g., Fleeson & Gallagher, influenced by dispositional affectivity or the stable
2009). More importantly, it is sometimes unclear aspects of transient mood states. Specifically, to the
to what extent the scales that are revised for use in extent that each report of conscientious behavior is
ESM reflect the structure and validity of the original, influenced consistently by negative or positive mood,
trait-level construct. That is, if a new scale is used, the aggregation of these experienced personality
often the psychometric properties of the scale are states will reflect elements of trait affectivity. Indices
defended only by the inclusion of coefficient alpha, of the evaluated self might also be influenced these
and it is unclear even whether this alpha reflects factors, but the nature of this influence will be less
the nested nature of the observations (i.e., internal systematic, as one’s current or dispositional mood is
consistency based on total variance vs. just within- only influencing a single response (i.e., to the global
person variance). Although it is perhaps true that self-report of personality). To the extent that moods
undertaking and reporting a separate validation vary, this influence will be less systematically tied to
effort for considerably shortened scales is unlikely affective dispositions.
More substantively, the difference between
to be seen as a contribution worthy of publication
assessments of the experienced self and the
in our more prestigious journals, one should at
evaluated self could reveal widely varying effects on
a minimum report the results of an appropriate
outcomes of interest to a wide array of disciplines.
multilevel confirmatory factor model along with a
Although the specification of a theory to determine
comparison to published models of trait measures.
the nature and magnitude of such differences is
Conceptual considerations.   Putting aside issues beyond the purview of this chapter, an existing
related to statistical modeling, we’d like to return and well-documented distinction in organizational
to the potential of this analysis to increase our psychology might provide a good starting point.
understanding of personality and other variables Specifically, as discussed by Weiss and Cropanzano
that have stable interindividual components (1996), affect-driven and judgment-driven behaviors
(e.g., job or life satisfaction). Assuming that in response to particular affective events seem to
reasonably reliable and valid measures can be fit as differential outcomes of the experienced and
created for assessing episodic or statelike analogues, evaluated self, respectively. Affect-driven behaviors
we suggest that whereas standard measures of are work-relevant actions that are determined
personality or other similarly stable constructs primarily by immediate affective states, which more

322
Looking Within

closely correspond to the experienced self. However, Bakker, A. B., Demerouti, E., Oerlemans, W., & Sonnentag, S.
judgment-driven behaviors likely take more time to (2013). Workaholism and daily recovery: A day
reconstruction study of leisure activities. Journal of
develop (i.e., may unfold at a differential rate) and Organizational Behavior, 34, 87–107. http://dx.doi.org/
are a function of work attitudes generated by the 10.1002/job.1796
aforementioned affective reactions and accumulated Bakker, A. B., & Xanthopoulou, D. (2009). The crossover
(and often more stable) beliefs. As stated by Weiss of daily work engagement: Test of an actor–partner
and Cropanzano (1996), judgment-driven behaviors interdependence model. Journal of Applied
Psychology, 94, 1562–1571. http://dx.doi.org/
“are the consequences of decision processes where 10.1037/a0017525
one’s evaluation of one’s job is part of the decision
Barker, R. G. (1968). Ecological psychology: Concepts
matrix” (p. 13). Although affective events theory and methods for studying the environment of human
(AET) represents only one conceptual basis on which behavior. Stanford, CA: Stanford University Press.
one might expect differences between the experienced Barnes, C., Lucianetti, L., Bhave, D., & Christian, M.
and evaluated selves, it is quite possible that many (2015). “You wouldn’t like me when I’m sleepy”:
Copyright American Psychological Association. Not for further distribution.

such differences exist across the psychological and Leaders’ sleep, daily abusive supervision, and
work unit engagement. Academy of Management
organizational sciences. The analysis of within-person Journal, 58, 1419–1437. http://dx.doi.org/10.5465/
processes and their aggregation to the between- amj.2013.1063
person level makes it clear that researchers have, up Beal, D. J. (2015). ESM 2.0: State of the art and
to now, been firmly rooted in only the evaluated self. future potential of experience sampling methods
We look forward to discovering the extent to which in organizational research. Annual Review of
Organizational Psychology and Organizational
a consideration of both perspectives can inform our Behavior, 2, 383–407. http://dx.doi.org/10.1146/
theories and our findings. annurev-orgpsych-032414-111335
Beal, D. J., & Dawson, J. F. (2007). On the use of
Likert-type scales in multilevel data influence on
CONCLUSION aggregate variables. Organizational Research
Over the course of this chapter, we have provided Methods, 10, 657–672. http://dx.doi.org/10.1177/
1094428106295492
an integrated overview of several intraindividual
Beal, D. J., & Ghandour, L. (2011). Stability, change,
methodologies as a means to illustrate not only what and the stability of change in daily workplace affect.
different types of empirical questions they address but Journal of Organizational Behavior, 32, 526–546.
also how they can be utilized in tandem. Specifically, http://dx.doi.org/10.1002/job.713
by using several intraindividual methods to delve into Beal, D. J., Trougakos, J. P., Weiss, H. M., & Dalal, R. S.
a phenomenon of interest, we posit that researchers (2013). Affect spin and the emotion regulation
process at work. Journal of Applied Psychology, 98,
can gain a more complete understanding of the lived 593–605. http://dx.doi.org/10.1037/a0032559
experience of employees on a day-to-day basis and
Beal, D. J., Trougakos, J. P., Weiss, H. M., & Green, S. G.
how these day-to-day experiences may culminate (2006). Episodic processes in emotional labor:
in meaningful, between-person characteristics. As Perceptions of affective delivery and regulation
scholars continue to increasingly employee these strategies. Journal of Applied Psychology, 91, 1053–1065.
http://dx.doi.org/10.1037/0021-9010.91.5.1053
methods, we hope that they consider using them
conjointly and in a more programmatic manner. Beal, D. J., & Weiss, H. M. (2003). Methods of ecological
momentary assessment in organizational research.
Organizational Research Methods, 6, 440–464.
References http://dx.doi.org/10.1177/1094428103257361
Andersson, L. M., & Pearson, C. M. (1999). Tit for tat? Beal, D. J., Weiss, H. M., Barros, E., & MacDermid, S. M.
The spiraling effect of incivility in the workplace. (2005). An episodic process model of affective
Academy of Management Journal, 24, 452–471. influences on performance. Journal of Applied
Psychology, 90, 1054–1068. http://dx.doi.org/
Aronson, E., & Carlsmith, J. M. (1968). Experimentation in
10.1037/0021-9010.90.6.1054
social psychology. In G. Lindzey & E. Aronson (Eds.),
Handbook of social psychology (Vol. 2, pp. 1–79). Benedetti, A. A., Diefendorff, J. M., Gabriel, A. S., &
Reading, MA: Addison-Wesley. Chandler, M. M. (2015). The effects of intrinsic and

323
Beal and Gabriel

extrinsic sources of motivation on well-being depend Csikszentmihalyi, M., Larson, R., & Prescott, S. (1977).
on time of day: The moderating effects of workday The ecology of adolescent activity and experience.
accumulation. Journal of Vocational Behavior, 88, Journal of Youth and Adolescence, 6, 281–294.
38–46. http://dx.doi.org/10.1016/j.jvb.2015.02.009 http://dx.doi.org/10.1007/BF02138940
Berg, K. C., Crosby, R. D., Cao, L., Peterson, C. B., Engel, Diener, E., & Tay, L. (2014). Review of the day
S. G., Mitchell, J. E., & Wonderlich, S. A. (2013). reconstruction method (DRM). Social Indicators
Facets of negative affect prior to and following Research, 116, 255–267. http://dx.doi.org/10.1007/
binge-only, purge-only, and binge/purge events in s11205-013-0279-x
women with bulimia nervosa. Journal of Abnormal Dimotakis, N., Ilies, R., & Judge, T. A. (2013).
Psychology, 122, 111–118. http://dx.doi.org/10.1037/ Experience sampling methodology. In J. M. Cortina
a0029703 & R. S. Landis (Eds.), Modern research methods for
Blau, P. M. (1964). Exchange and power in social life. the study of behavior in organizations (pp. 319–348).
New York, NY: Wiley. New York, NY: Routledge.
Bliese, P., Chan, D., & Ployhart, R. E. (2007). Multilevel Dimotakis, N., Scott, B. A., & Koopman, J. (2011). An
methods: Future directions in measurement, experience sampling investigation of workplace
Copyright American Psychological Association. Not for further distribution.

longitudinal analyses, and nonnormal outcomes. interactions, affective states, and employee well-
Organizational Research Methods, 10, 551–563. being. Journal of Organizational Behavior, 32,
http://dx.doi.org/10.1177/1094428107301102 572–588. http://dx.doi.org/10.1002/job.722
Bolger, N., Davis, A., & Rafaeli, E. (2003). Diary Edwards, J. R. (2008). To prosper organizational psychology
methods: Capturing life as it is lived. Annual Review should . . . overcome methodological barriers to
of Psychology, 54, 579–616. http://dx.doi.org/ progress. Journal of Organizational Behavior, 29,
10.1146/annurev.psych.54.101601.145030 469–491. http://dx.doi.org/10.1002/job.529
Butts, M. M., Becker, W. J., & Boswell, W. R. (2015). Hot Eisenberger, N. I., Taylor, S. E., Gable, S. L., Hilmert,
buttons and time sinks: The effects of electronic C. J., & Lieberman, M. D. (2007). Neural pathways
communication during nonwork time on emotions link social support to attenuated neuroendocrine
and work–nonwork conflict. Academy of Management stress responses. Neuroimage, 35, 1601–1612.
Journal, 58, 763–788. http://dx.doi.org/10.5465/ Fisher, C. D., & To, M. L. (2012). Using experience
amj.2014.0170 sampling methodology in organizational behavior.
Journal of Organizational Behavior, 33, 865–877.
Chan, D. (1998). Functional relations among constructs
http://dx.doi.org/10.1002/job.1803
in the same content domain at different levels of
analysis: A typology of composition models. Journal Fleeson, W. (2001). Toward a structure- and process-
of Applied Psychology, 83, 234–246. http://dx.doi.org/ integrated view of personality: Traits as density
10.1037/0021-9010.83.2.234 distribution of states. Journal of Personality and Social
Psychology, 80, 1011–1027. http://dx.doi.org/
Chen, G., Bliese, P., & Mathieu, J. (2005). Conceptual 10.1037/0022-3514.80.6.1011
framework and statistical procedures for delineating
and testing multilevel theories of homology. Fleeson, W. (2004). Moving personality beyond the
Organizational Research Methods, 8, 375–409. person-situation debate: The challenge and the
http://dx.doi.org/10.1177/1094428105280056 opportunity of within-person variability. Current
Directions in Psychological Science, 13, 83–87.
Clarivate Analytics. (2018, April 20). Retrieved from the http://dx.doi.org/10.1111/j.0963-7214.2004.00280.x
University of Arizona’s subscription-based Web of
Science database. Fleeson, W., & Gallagher, P. (2009). The implications
of Big Five standing for the distribution of trait
Cropanzano, R., & Mitchell, M. S. (2005). Social manifestation in behavior: Fifteen experience-
exchange theory: An interdisciplinary sampling studies and a meta-analysis. Journal of
review. Journal of Management, 31, 874–900. Personality and Social Psychology, 97, 1097–1114.
http://dx.doi.org/10.1177/0149206305279602 http://dx.doi.org/10.1037/a0016786
Csikszentmihalyi, M., & Graef, R. (1980). The experience Foo, M.-D., Uy, M. A., & Baron, R. A. (2009). How do
of freedom in daily life. American Journal of Community feelings influence effort? An empirical study of
Psychology, 8, 401–414. http://dx.doi.org/10.1007/ entrepreneurs’ affect and venture effort. Journal of
BF00912853 Applied Psychology, 94, 1086–1094. http://dx.doi.org/
Csikszentmihalyi, M., & Larson, R. (1987). Validity 10.1037/a0015599
and reliability of the experience-sampling Fredrickson, B. L. (2000). Extracting meaning from past
method. Journal of Nervous and Mental Disease, affective experiences: The importance of peaks, ends,
175, 526–536. http://dx.doi.org/10.1097/ and specific emotions. Cognition and Emotion, 14,
00005053-198709000-00004 577–606. http://dx.doi.org/10.1080/026999300402808

324
Looking Within

Fredrickson, B. L., & Kahneman, D. (1993). Duration Groth, M., & Grandey, A. (2012). From bad to worse:
neglect in retrospective evaluations of affective Negative exchange spirals in employee–customer
episodes. Journal of Personality and Social service interactions. Organizational Psychology
Psy­chology, 65, 45–55. http://dx.doi.org/10.1037/ Review, 2, 208–233. http://dx.doi.org/10.1177/
0022-3514.65.1.45 2041386612441735
Gabriel, A. S., & Diefendorff, J. M. (2015). Emotional Ilies, R., & Judge, T. A. (2002). Understanding the
labor dynamics: A momentary approach. Academy of dyna­mic relationships among personality, mood,
Management Journal, 58, 1804–1825. http://dx.doi.org/ and job satisfaction: A field experience sampling
10.5465/amj.2013.1135 study. Organizational Behavior and Human Decision
Processes, 89, 1119–1139. http://dx.doi.org/
Gabriel, A. S., Diefendorff, J. M., Bennett, A. A., &
10.1016/S0749-5978(02)00018-3
Sloan, M. D. (2017). It’s about time: The promise of
continuous rating assessments for the organizational Jones, K. P. (2017). To tell or not to tell? Examining the
sciences. Organizational Research Methods, 20, 32–60. role of discrimination in the pregnancy disclosure
http://dx.doi.org/10.1177/1094428116673721 process at work. Journal of Occupational Health
Psychology, 22, 239–250. http://dx.doi.org/
Gabriel, A. S., Diefendorff, J. M., Chandler, M. M.,
Copyright American Psychological Association. Not for further distribution.

10.1037/ocp0000030
Moran, C. M., & Greguras, G. J. (2014). The dynamic
relation­ships of work affect and job satisfaction with Judge, T. A., & Ilies, R. (2004). Affect and job satis­fac­
perceptions of fit. Personnel Psychology, 67, 389–420. tion: A study of their relationship at work and at
http://dx.doi.org/10.1111/peps.12042 home. Journal of Applied Psychology, 89, 661–673.
http://dx.doi.org/10.1037/0021-9010.89.4.661
Gabriel, A. S., Diefendorff, J. M., & Erickson, R. J. (2011).
The relations of daily task accomplishment satisfaction Judge, T. A., Scott, B. A., & Ilies, R. (2006). Hostility, job
with changes in affect: A multilevel study in nurses. attitudes, and workplace deviance: Test of a multilevel
Journal of Applied Psychology, 96, 1095–1104. model. Journal of Applied Psychology, 91, 126–138.
http://dx.doi.org/10.1037/a0023937 http://dx.doi.org/10.1037/0021-9010.91.1.126
George, J. M., & Jones, G. R. (2000). The role of time Judge, T. A., Woolf, E. F., & Hurst, C. (2009). Is
in theory and theory building. Journal of Manage­ emotional labor more difficult for some than for
ment, 26, 657–684. http://dx.doi.org/10.1177/ others? A multilevel, experience-sampling study.
014920630002600404 Personnel Psychology, 62, 57–88. http://dx.doi.
org/10.1111/j.1744-6570.2008.01129.x
Glaser, B. G., & Strauss, A. L. (1967). The discovery of
grounded theory: Strategies for qualitative research. Kahneman, D., Krueger, A. B., Schkade, D. A., Schwarz, N.,
Hawthorne, NY: Aldine de Gruyter. & Stone, A. A. (2004). A survey method for
characterizing daily life experience: The day
Glomb, T. M., Beal, D. J., Yang, T., & Bhave, D. P. (2014, reconstruction method. Science, 306, 1776–1780.
May). Staying power: Emotional inertia as a moderator http://dx.doi.org/10.1126/science.1103572
of event-affect relationships. Paper presented at the
29th Annual Conference of the Society of Industrial Kahneman, D., & Riis, J. (2005). Living and thinking
and Organizational Psychology, Honolulu, HI. about it: Two perspectives on life. In N. Baylis,
F. A. Huppert, & B. Keverne (Eds.), The science of
Gobo, G. (2004). Sampling, representativeness, and well-being (pp. 284–304). New York, NY: Oxford
generalizability. In C. Seale, G. Gobo, J. F. Gubrium, Press. http://dx.doi.org/10.1093/acprof:oso/
& D. Silverman (Eds.), Qualitative health research 9780198567523.003.0011
(pp. 435–455). Thousand Oaks, CA: Sage.
Koopman, J., Lanaj, K., & Scott, B. A. (2016). Integrating
Gottman, J., Notarius, C., Markman, H., Bank, S., Yoppi, B., the bright and dark sides of OCB: A daily investigation
& Rubin, M. E. (1976). Behavior exchange theory of the benefits and costs of helping others. Academy
and marital decision making. Journal of Personality of Management Journal, 59, 414–435. http://dx.doi.org/
and Social Psychology, 34, 14–23. http://dx.doi.org/ 10.5465/amj.2014.0262
10.1037/0022-3514.34.1.14
Kozlowski, S. W. J., & Klein, K. J. (2000). A multilevel
Gottman, J. M., & Levenson, R. W. (1985). A valid approach to theory and research in organizations:
procedure for obtaining self-report of affect in Contextual, temporal, and emergent processes. In
marital interaction. Journal of Consulting and Clinical K. Klein & S. W. J. Kowzlowski (Eds.), Multilevel
Psychology, 53, 151–160. http://dx.doi.org/ theory, research, and methods in organizations: Founda­
10.1037/0022-006X.53.2.151 tions, extensions, and new directions (pp. 3–90).
Grandey, A. A. (2000). Emotion regulation in the work­ San Francisco, CA: Jossey-Bass.
place: A new way to conceptualize emotional labor. Kuppens, P., Oravecz, Z., & Tuerlinckx, F. (2010). Feelings
Journal of Occupational Health Psychology, 5, 95–110. change: Accounting for individual differences in the
http://dx.doi.org/10.1037/1076-8998.5.1.95 temporal dynamics of affect. Journal of Personality and

325
Beal and Gabriel

Social Psychology, 99, 1042–1060. http://dx.doi.org/ Monge, P. R. (1990). Theoretical and analytical issues
10.1037/a0020962 in studying organizational processes. Organization
Science, 1, 406–430. http://dx.doi.org/10.1287/
Laurenceau, J.-P., Barrett, L. F., & Rovine, M. J. (2005).
orsc.1.4.406
The interpersonal process model of intimacy in
marriage: A daily-diary and multilevel modeling Morgeson, F. P., Mitchell, T. R., & Liu, D. (2015). Event
approach. Journal of Family Psychology, 19, 314–323. system theory: An event-oriented approach to the
http://dx.doi.org/10.1037/0893-3200.19.2.314 organizational science. Academy of Management Review,
40, 515–537. http://dx.doi.org/10.5465/amr.2012.0099
Leslie, L. M., King, E. B., Bradley, J. C., & Hebl, M. R.
(2008). Triangulation across methodologies: Naidoo, L. J., & Lord, R. G. (2008). Speech imagery
All signs point to persistent stereotyping and and perceptions of charisma: The mediating
discrimination in organizations. Industrial and role of positive affect. The Leadership Quarterly,
Organizational Psychology: Perspectives on Science 19, 283–296. http://dx.doi.org/10.1016/
and Practice, 1, 399–404. http://dx.doi.org/10.1111/ j.leaqua.2008.03.010
j.1754-9434.2008.00073.x Nishina, A., & Juvonen, J. (2005). Daily reports of
witnessing and experiencing peer harassment in
Copyright American Psychological Association. Not for further distribution.

Levenson, R. W., & Gottman, J. M. (1983). Marital


interaction: Physiological linkage and affective middle school. Child Development, 76, 435–450.
exchange. Journal of Personality and Social http://dx.doi.org/10.1111/j.1467-8624.2005.00855.x
Psy­chology, 45, 587–597. http://dx.doi.org/ Oerlemans, W. G., & Bakker, A. B. (2014). Burnout and
10.1037/0022-3514.45.3.587 daily recovery: A day reconstruction study. Journal
of Occupational Health Psychology, 19, 303–314.
Matta, F. K., Erol-Korkmaz, H. T., Johnson, R. E., &
http://dx.doi.org/10.1037/a0036904
Biçaksiz, P. (2014). Significant work events and
counterproductive behavior: The role of fairness, Ohly, S., Sonnentag, S., Niessen, C., & Zapf, D. (2010).
emotions, and emotion regulation. Journal of Organi­ Diary studies in organizational research: An intro­
zational Behavior, 35, 920–944. http://dx.doi.org/ duction and some practical recommendations. Journal
10.1002/job.1934 of Personnel Psychology, 9, 79–93. http://dx.doi.org/
10.1027/1866-5888/a000009
Matta, F. K., Scott, B. A., Colquitt, J. A., Koopman, J.,
& Passantino, L. G. (2017). Is consistently unfair Ployhart, R. E., & Vandenberg, R. J. (2010). Longitudinal
better than sporadically fair? An investigation of research: The theory, design, and analysis of change.
justice variability and stress. Academy of Management Journal of Management, 36, 94–120. http://dx.doi.org/
Journal, 60, 743–770. http://dx.doi.org/10.5465/ 10.1177/0149206309352110
amj.2014.0455 Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff,
Mauss, I. B., Levenson, R. W., McCarter, L., Wilhelm, N. P. (2003). Common method biases in behavioral
F. H., & Gross, J. J. (2005). The tie that binds? research: A critical review of the literature and
Coherence among emotion experience, behavior, and recommended remedies. Journal of Applied
Psychology, 88, 879–903. http://dx.doi.org/
physiology. Emotion, 5, 175–190. http://dx.doi.org/
10.1037/0021-9010.88.5.879
10.1037/1528-3542.5.2.175
Robinson, M. D., & Clore, G. L. (2002). Belief and feeling:
Mauss, I. B., Shallcross, A. J., Troy, A. S., John, O. P.,
Evidence for an accessibility model of emotional
Ferrer, E., Wilhelm, F. H., & Gross, J. J. (2011). self-report. Psychological Bulletin, 128, 934–960.
Don’t hide your happiness! Positive emotion http://dx.doi.org/10.1037/0033-2909.128.6.934
dissociation, social connectedness, and psychological
functioning. Journal of Personality and Social Rodell, J. B., & Judge, T. A. (2009). Can “good” stressors
Psychology, 100, 738–748. http://dx.doi.org/ spark “bad” behaviors? The mediating role of emotions
10.1037/a0022410 in links of challenge and hindrance stressors with
citizenship and counterproductive behaviors. Journal of
Methot, J. R., Lepak, D., Shipp, A. J., & Boswell, W. R. Applied Psychology, 94, 1438–1451. http://dx.doi.org/
(2017). Good citizen interrupted: Calibrating a 10.1037/a0016752
temporal theory of citizenship behavior. Academy of
Management Review, 42, 10–31. http://dx.doi.org/ Rosen, C. C., Koopman, J., Gabriel, A. S., & Johnson,
10.5465/amr.2014.0415 R. E. (2016). Who strikes back? A daily investigation
of when and why incivility begets incivility. Journal of
Miron-Shatz, T. (2009). Evaluating multiepisode events: Applied Psychology, 101, 1620–1634. http://dx.doi.org/
Boundary conditions for the peak-end rule. Emotion, 10.1037/apl0000140
9, 206–213. http://dx.doi.org/10.1037/a0015295
Ruef, A. M., & Levenson, R. W. (2007). Continuous
Mitchell, T. R., & James, L. R. (2001). Building better measurement of emotion. In J. A. Coan & J. B. Allen
theory: Time and the specification of when things (Eds.), Handbook of emotion elicitation and assessment
happen. Academy of Management Review, 26, 530–547. (pp. 286–297). New York, NY: Oxford University Press.

326
Looking Within

Rupp, D. E., & Spencer, S. (2006). When customers lash Tracy, J. L., & Robins, R. W. (2007). The psychological
out: The effects of customer interactional injustice structure of pride: A tale of two facets. Journal of
on emotional labor and the mediating role of discrete Personality and Social Psychology, 92, 506–525.
emotions. Journal of Applied Psychology, 91, 971–978. http://dx.doi.org/10.1037/0022-3514.92.3.506
http://dx.doi.org/10.1037/0021-9010.91.4.971 Tracy, J. L., Shariff, A. F., & Cheng, J. T. (2010). A
Russell, J. A., & Feldman Barrett, L. (1999). Core affect, naturalist’s view of pride. Emotion Review, 2, 163–177.
prototypical emotional episodes, and other things http://dx.doi.org/10.1177/1754073909354627
called emotion: Dissecting the elephant. Journal Trougakos, J. P., Beal, D. J., Cheng, B. H., Hideg, I., &
of Personality and Social Psychology, 76, 805–819. Zweig, D. (2015). Too drained to help: A resource
http://dx.doi.org/10.1037/0022-3514.76.5.805 depletion perspective on daily interpersonal
Sander, D., Grandjean, D., & Scherer, K. R. (2005). citizenship behaviors. Journal of Applied Psychology,
A systems approach to appraisal mechanisms 100, 227–236. http://dx.doi.org/10.1037/a0038082
in emotion. Neural networks, 18, 317–352. Wagner, D. T., Barnes, C. M., & Scott, B. A. (2014).
http://dx.doi.org/10.1016/j.neunet.2005.03.001 Driving it home: How workplace emotional labor
Scott, B. A., & Barnes, C. M. (2011). A multilevel harms employee home life. Personnel Psychology, 67,
Copyright American Psychological Association. Not for further distribution.

investiga­tion of emotional labor, affect, with­ 487–516. http://dx.doi.org/10.1111/peps.12044


drawal, and gender. Academy of Management Wang, L. P., Hamaker, E., & Bergeman, C. S. (2012).
Journal, 54, 116–136. http://dx.doi.org/10.5465/ Investigating inter-individual differences in short-term
AMJ.2011.59215086 intraindividual variability. Psychological Methods, 17,
Scott, B. A., Barnes, C. M., & Wagner, D. T. (2012). 567–581. http://dx.doi.org/10.1037/a0029317
Chameleonic or consistent? A multilevel investigation Wang, M., Liu, S., Liao, H., Gong, Y., Kammeyer-Mueller, J.,
of emotional labor variability and self-monitoring. & Shi, J. (2013). Can’t get it out of my mind:
Academy of Management Journal, 55, 905–926. Employee rumination after customer mistreatment
http://dx.doi.org/10.5465/amj.2010.1050 and negative mood in the next morning. Journal of
Siewert, K., Antoniw, K., Kubiak, T., & Weber, H. (2011). Applied Psychology, 98, 989–1004. http://dx.doi.org/
The more the better? The relationship between 10.1037/a0033656
mismatches in social support and subjective Weiss, H. M., Beal, D. J., Lucy, S. L., & MacDermid, S. M.
well-being in daily life. Journal of Health (2004). Constructing EMA studies with PMAT: The
Psychology, 16, 621–631. http://dx.doi.org/ Purdue momentary assessment tool user’s manual.
10.1177/1359105310385366 West Lafayette, IN: Purdue University.
Sonnentag, S., Binnewies, C., & Mojza, E. J. (2008). Weiss, H. M., & Cropanzano, R. (1996). Affective events
“Did you have a nice evening?” A day-level study theory: A theoretical discussion of the structure,
on recovery experiences, sleep, and affect. Journal of causes, and consequences of affective experiences at
Applied Psychology, 93, 674–684. http://dx.doi.org/ work. Research in Organizational Behavior, 18, 1–74.
10.1037/0021-9010.93.3.674
Weiss, H. M., & Rupp, D. E. (2011). Experiencing work:
Sonnentag, S., & Zijlstra, F. R. H. (2006). Job character­ An essay on a person-centric work psychology.
istics and off-job activities as predictors of need for Industrial and Organizational Psychology: Perspectives
recovery, well-being, and fatigue. Journal of Applied on Science and Practice, 4, 83–97. http://dx.doi.org/
Psychology, 91, 330–350. http://dx.doi.org/ 10.1111/j.1754-9434.2010.01302.x
10.1037/0021-9010.91.2.330
Yang, J., & Diefendorff, J. M. (2009). The relations of
Swallow, K. M., Zacks, J. M., & Abrams, R. A. (2009). daily counterproductive workplace behavior with
Event boundaries in perception affect memory emotions, situational antecedents, and personality
encoding and updating. Journal of Experimental moderators: A diary study in Hong Kong. Personnel
Psychology: General, 138, 236–257. http://dx.doi.org/ Psychology, 62, 259–295. http://dx.doi.org/10.1111/
10.1037/a0015631 j.1744-6570.2009.01138.x

327

You might also like