EMOTION-REGULATION IN SERVICE ENCOUNTER: COPING WITH BEHAVIORS OF
OTHER CUSTOMERS
Li Miao
Purdue University
lmiao@purdue.edu
Abstract
Hospitality consumption, such as dining in a restaurant or taking a cruise vacation, often involves
sharing the service environment with other customers. While the presence of other consumers in the service
environment has long been recognized in the hospitality literature, research in this area has traditionally focused
on customer-servicescape interface and customer-provider dynamics. The impact of other customers on
consumption experience has received scant attention (Tombs & McColl-Kennedy, 2003). The lack of
systematic research on the impact of “other customers” is surprising given that in many consumption situations,
other customers are an integral part of the experience and can have a significant impact on a focal customer’s
overall experience (Grove & Fisk, 1997).
The present research examined how the behavior of other customers affects individuals’ emotional
experiences and how the individuals regulate their emotional states to cope with the behavior of others in service
encounters. The current research consists of two studies. Study 1 examined customers’ emotional responses to
the behavior of other customers. Study 2 investigated the relationships among emotional responses, coping
strategies and encounter satisfaction.
Theoretical Background Emotional Responses to the Behavior of Other Customers
The extant literature has referred to other customers as “audience” (Grove & Fisk, 1983), “participants”
(Booms & Bitner, 1981), “Customer B” (Langeard et al., 1981) or “co-actors” (Aubert-Gamet, 1999) at service
encounters. The presence of other customers weaves a complex social and psychological landscape in which
services are consumed. Considerable evidence suggested that an individual’s experience is affected by “the
actual, imagined, or implied presence of others” (Allport 1985, p.3). Emotions are arguably the most
fundamental dimension of interpersonal behaviors (Forgas, 2000; Zajonc, 1980). Emotions are defined as
“intense, short-lived and usually have a definite cause and clear cognitive content” (Forgas, 1992, p.230). The
interpersonal perspective of emotions posits that emotions are socially constituted over the course of
interpersonal encounters (e.g., Averill, 1985; Parkinson, 1995). Kemper (1978) suggested that the most
important objects in anyone’s environment are other people, what they say, and what they do. Many emotions
are explicitly directed at other people and arise out of interactions with them. Parkinson (1995) went so far as to
argue that other people supply the most important class of emotion-inducing events. Anger, for example, does
not usually make sense without someone else whom to be angry at. The valence of others’ behaviors (positive
or negative) determines whether emotional responses are positive, neutral or negative (Graesser et al. 1979;
Graesser et al. 1980; Maki 1990). To empirically test this proposition in the context of service encounters, the
following hypothesis is proposed:
H1: Emotional responses to behavior of other customers differ by the valence of the behavior, such that
emotional responses in the positive valence group are the most positive, followed by those of the control group
and of the negative valence group respectively.
Emotional Displays
Hochschild (1983) suggested that consumers and service providers share a set of expectations in
emotional displays in a service encounter. The concept of display rules (Ekman, 1972) holds that “the
expression of one’s internal feeling state may be controlled and modified in a variety of ways, by presenting an
expression that minimizes, exaggerates, or masks the feeling state to suit the particular demands of the social
situation” (Friedman & Miller-Herringer, 1991, p.766). Display rules are thought to depend upon shared beliefs
about appropriateness of expressions (Ekman, 1973). In the consumer literature, display rules are investigated
mainly in consumer-provider interactions (e.g., Rafaeli & Sutton, 1989, 1990; Sutton, 1991). The expectations
for service providers’ emotional displays are a function of societal norms, occupational norms and
organizational norms (Rafaeli & Sutton, 1989). As service encounters are also social encounters, it can be
inferred that consumers also hold a set of expectations about emotional displays in a particular service encounter.
Extant literature is scant on emotional displays in consumer-to-consumer interactions at service encounters. To
that end, this study explores emotional expressions toward behaviors of other consumers.
It is widely agreed that emotional displays evolved to serve social functions and such emotional
displays are particularly susceptible to social influences (Buck et al., 1992). Presence of others complicates
emotional displays especially when others function as eliciting stimuli themselves (Buck, 1990). In fact,
displayed emotions are a form of communication (Mattila & Enz, 2002; Rafaeli & Sutton, 1990). Interpersonal
view of emotions (Parkinson, 1995) suggests that consumers’ emotional states are communicated through
emotional displays to other consumers who function as eliciting stimuli. Emotional displays convey important
information about how the consumer will ultimately assess a service encounter.
The motivational view (Berger et al., 1981; Blank, 1980) suggests that people suppress the expression
of socially undesirable behaviors to avoid social disapproval. The degree of inhibition of emotional expressions
depends on the social consequences of the expressions (Friedman & Miller-Herringer, 1991). Other consumers’
negative behaviors mirror a situation where responses to such behaviors have potential for causing a person
embarrassment or loss of esteem of others. In this case, inhibition of undesirable behaviors is likely to occur. In
essence, the proposition is that the very presence of other consumers changes the way a focal consumer
responds to the behaviors demonstrated by other consumers, especially when such behaviors are negative.
Inhibition of emotional expressions is likely to cause a mismatch between the felt emotions and expressed
emotions, an emotional state labeled as emotive dissonance (Kruml & Geddes, 2000). Hence, it is proposed:
H2: Emotive dissonance experienced by individuals differs by the valence of the behavior of other customers,
such that the level of emotive dissonance experienced by individuals in the negative valence group is the
greatest, followed by that of the control group and of the positive valence group respectively.
Study 1: Emotional Responses to the Behavior of Other Customers
Subjects, Design, Procedure, and Experimental Conditions
A total of 137 undergraduate students at a major university in the Northeastern region of the United
States participated in Study 1 in exchange for extra credit. Fifty four percents of the participants were female
and the average age was 21 years. A control group was added to provide baseline measures for the dependent
variables. The control group represents a condition with no manipulation of other customers’ behavior. The
condition is described as “typical” or “as expected”. Participants were randomly assigned to the five conditions
and were exposed to a hypothetical consumption episode in a restaurant setting.
Independent Variables
Manipulation of Other Customers’ Behaviors. The behaviors of other customers (negative vs. positive) were
manipulated by describing other customers’ specific behaviors in the service encounter scenarios. Participants
read either a positively framed or a negatively framed scenario. The choice of specific behaviors was based on
the results of the three pilot studies. The pilot studies also verified that the magnitudes of the positive and
negative behaviors were equivocal. In the positive behavior condition, the scenario described that the children
in the service setting were courteous and their parents made sure that the children were on their best behavior.
Conversely, in the negative behavior condition, the scenario described the children as disruptive, and their
parents made no effort to keep the children under control. The selection of other customers’ behaviors was
based on a pilot study (n =73) with a sample of undergraduate students recruited from the same population as
Study 1.
Dependent Variables
Emotional Responses. To assess emotional responses, participants completed an alphabetized seven-item
emotion scale, based largely on Richins’s (1997) work, with relevant items added from Sedikides and Gaertner
(2001) and Oliver (2000). These items consisted of emotional terms such as “pleased”, “delighted”, “happy”,
“annoyed”, “angry”, “frustrated” and “irritated”. Participants were asked to indicate the extent to which each
item reflects how they feel when they imagine themselves in the service encounter described in the scenario.
All responses were on a seven-point scale anchored at 1= not at all and 7 = very much. This seven-item scale
was used to derive an index for positive and negative emotional responses (three positive items, α = 0.96, four
negative items, α = 0.97). A positivity index was constructed by averaging the positive and negative emotional
responses separately and then subtracting the later from the former for each participant (Labroo & Ramanathan,
2007).
Emotive dissonance. Emotive dissonance was measured via a two-item scale (Kruml & Geddes, 2000). The
two items measured on a seven-point scale were “If I were in the situation described in the scenario, I would
show the same feelings that I feel inside” and “If I were in the situation described in the scenario, the emotions I
show would match what I truly feel”. Ratings on the two items (r = 0.80) were recoded so that greater scores
represent greater levels of emotive dissonance.
Manipulation checks were also included in the research protocol. Realism of the scenario was
measured with the item “how realistic is the scenario described at the beginning of the questionnaire?” (1= very
unrealistic and 7= very realistic). Ease of use was measured with the item “how easy is it for you to imagine
yourself in the scenario?” (1= very difficult and 7= very easy).
Results
Manipulation Checks. The scenario manipulations were first checked for their realism (M’s ≥ 5.28, SD’s ≤ .32)
and ease of use (M’s ≥ 5.53, SD’s ≤ .24). Furthermore, as expected, participants in the negative behavior
condition rated the behaviors more negatively than those in the positive behavior condition (M’s = -1.60 vs. 2.10;
t (139) = -21.22, p < 0.001) and both were significantly different from the midpoint (t (70) = -13.28 and t (70) =
16.66, p’s < 0.001). Taken together, the manipulation of other customers’ behaviors was successful.
A MANOVA analysis was performed on the set of the dependent variables with behaviors of other
customers as the between-subjects factor. The MANOVA results showed that the main effect of the behavior of
other customers (F(4,320) = 104.79, p < 0.001, Wilks’s λ = .18) was highly significant. In the following section,
the MANOVA results on each dependent variable are reported in detail. Table 1 presents the means and
standard errors of the dependent variable measures.
Table 1
Means and Standard Errors of Dependent Measures
Dependent Measures
Negative Behavior Group
Control Group
Positive Behavior Group
M
SE
M
SE
M
SE
Emotional responses
Emotive dissonance
-3.01
4.43
.22
.16
4.60
2.84
.39
.28
3.71
4.09
.22
.16
Emotional Responses. The MANOVA results showed a significant effect of behaviors of other customers on
emotional responses (F (2, 161) = 273.79, p < 0.001). The Post-Hoc LSD multiple comparison results revealed
that emotional responses in the three groups (negative, control and positive) were all significantly different from
each other. As expected, emotional responses of the group exposed to the positive behavior and of the control
group were significantly more positive than those exposed to the negative behavior (Mpositive = 3.71, SE =.22 vs.
Mnegative = -3.01, SE =.22; p < .001; Mcontrol = 4.60, SE = .39 vs. Mnegative = -3.01, SE = .22; p < .001). However,
an interesting and unexpected pattern of emotional responses between the positive group and the control group
emerged from the results. The results showed that emotional responses in the control group (behavior of other
customers was described as “typical” and “as expected’) were significantly more positive than those in the
positive group (in which behavior of other customers was described as “positive” and “desirable”) (Mcontrol =
4.60, SE = .39 vs. Mpositive = 3.71, SE = .22 ; p < .05). Taken together, Hypothesis 1 was partially supported.
Figure 1 illustrates the pattern of the emotional responses among the three groups.
Figure 1
Emotional Responses and Emotive Dissonance among the Three Experimental Groups
Emotive Dissonance. The MANOVA results revealed a significant effect of behaviors of other customers on
emotive dissonance (F (2, 162) = 12.07, p < 0.001). The Post-Hoc LSD multiple comparison results showed
that, as expected, emotive dissonance experienced by the negative group was significantly greater that that by
the control group (Mnegative = 4.43, SE =.16 vs. Mcontrol = 2.84, SE = .28; p < .001). However, the level of
emotive dissonance experienced by the negative and the positive group was similar (Mnegative = 4.43, SE = .16 vs.
Mpositive = 4.09, SE = .16; p = .13). More interestingly, emotive dissonance experienced by the positive group (in
which behavior of other customers was described as “positive” and “desirable”) was greater than that by the
control group (behavior of other customers was described as “typical” and “as expected’) (Mpositive = 4.09, SE
= .16 vs. Mcontrol = 2.84, SE = .28; p < .001). Taken together, Hypothesis 2 was partially supported. Figure 1
illustrates the pattern of the emotive dissonance among the three groups.
Discussion
The results generally support the prediction that emotional responses to the behavior of other customers
are influenced by the valence of the behavior, such that favorable behaviors will elicit more positive emotional
responses. An unexpected finding from Study 1 is that emotional responses of the control group (in which the
behaviors of other customers were described as “typical” and “as expected”) are significantly more positive than
those of individuals exposed to the positive behaviors of other customers (in which the behaviors of other
customers were described as “positive” and “desirable”). The result alludes to the certain level of coping on the
part of customers, even with behaviors of other customers that are considered favorable. As dealing with others’
behaviors can be psychologically taxing, positivity of emotional responses is adversely affected. The pattern of
results in emotive dissonance further confirms the presence of underlying coping mechanisms when sharing
service environment with other customers. The results show that consumers experience a mismatch between
their internal felt emotions and external emotional displays even when the behavior of other customers and their
emotional responses to the behavior of others are positive. The subjective experience of emotive dissonance, or
the inconsistency between how one feels and how one expresses that particular feeling, reveals a certain level of
emotion regulation (managing one’s emotions) on the part of the consumer. To further understand how
consumers regulate other-customers-inflicted emotions, Study 2 was conducted to investigate how consumers
cope with behaviors of others and how their emotional responses and coping strategies affect encounter
satisfaction, the ultimate outcome of the service experience.
Study 2: Coping with the Behavior of Other Customers
Coping – the Behavioral Responses of Emotion Regulation
Emotions can be regulated through a variety of behavioral responses, often referred to as coping
(Garnefski & Spinhoven, 2011). When other consumers demonstrate negative behaviors in a shared service
environment, the service encounter can be a stressful event which requires a certain degree of coping (Folkman,
Lazarus, Dunkel-Schetter, DeLongis & Gruen, 1986). Coping is assessed as a response to the psychological and
environmental demands of stressful encounters (Folkman et al., 1986). It is defined as a person’s constantly
changing cognitive and behavioral efforts to manage specific external and/or internal demands that are appraised
as taxing or exceeding the person’s resources (Lazarus & Folkman, 1984). Coping involves multiple strategies
which fall into two general types of coping: problem-focused coping and emotion-focused coping. Problemfocused coping is aimed at doing something to alter the troubled person-environment relation causing the
distress, and emotion-focused coping is oriented toward regulating negative emotions (Carver, Scheier &
Weintraub, 1989; Folkman et al., 1986). Coping strategies are considered behavioral manifestations of emotion
regulation (Garnefski & Spinhoven, 2011).
According to COPE inventory (Carver et al., 1989), in each of the two broad categories of coping, there
are specific coping activities. For example, in the category of problem-focused coping, specific activities
include active coping planning, suppression of competing activities, restraint coping, seeking social support for
instrumental reasons and seeking social support for emotional reasons. In the category of emotion-focused
coping, specific activities include positive reinterpretation and growth, acceptance, turning to religion, venting
of emotions, denial, behavioral disengagement, mental disengagement and alcohol-drug disengagement. The
pattern of coping responses is complex and is determined by many variables (Carver et al., 1989). However, it
is generally believed that problem-focused coping tends to predominate when people are in encounters that are
considered as changeable, whereas emotion-focused coping tends to predominate in encounters that are thought
to be unchangeable (Folkman & Lazarus, 1980). In addition, variability in coping is also partially a function of
people’s judgments about what is at stake in specific stressful encounters and what they view as the options for
coping (Folkman et al., 1986). In stressful encounters in which minimal psychological and social stakes are
involved, people use significantly less confrontational problem-focused coping and are more likely to engage in
escape-avoidance emotion-focused coping (Folkman et al., 1986). In a typical service encounter, the outcome
of the service consumption is more contingent on customer-provider interactions than customer-to-customer
interactions. We predict consumers are more likely to use emotion-focused coping strategies to deal with the
behavior of other customers than the more confrontational problem-focused coping strategies. The hypothesis is
summarized below:
H3: In response to the negative behaviors of other customers at service encounters, an individual is more likely
to engage in emotion-focused coping than problem-focused coping.
Encounter Satisfaction
Consumer responses to service encounters have been defined predominantly in terms of encounter
satisfaction (Fisk, Brown & Bitner, 1993). Encounter satisfaction is typically assumed to vary along a
continuum from unfavorable (dissatisfied) to favorable (satisfied) (Price et al., 1995). Satisfaction judgments
are believed to represent a higher-order cumulative or global evaluation of the service encounter (Anderson &
Fornell, 2000). In this study, satisfaction refers to transaction-specific satisfaction. The transaction-based
conceptualization views satisfaction as a post-purchase evaluative judgment for a specific purchase occasion
(Anderson & Fornell, 1994). Consumer literature suggests that consumers’ emotional responses are moment-tomoment responses and such moment-to-moment responses can be integrated to overall judgment such as
encounter satisfaction (Baumgartner, Sujan & Padgett, 1997). Therefore, we predict that other-consumerelicited emotional responses and coping strategies will be integrated into encounter satisfaction. Positive
emotional responses to the behavior of others will favorably affect encounter satisfaction while coping, a
psychologically taxing process, will adversely affect encounter satisfaction. Furthermore, it is interesting to
explore whether emotional responses and coping strategies affect encounter satisfaction separately or the two
consumer responses interact with each other to jointly affect encounter satisfaction. As coping is a behavioral
response of emotion regulation and emotional responses can be a result of the regulation mechanism, we predict
that an interaction effect between the two may be present. We further predict that problem-focused coping
strategies can accentuate the effect of emotional responses on encounter satisfaction, because problem-focused
strategies are more confrontational and call for more efforts on the part of the consumers. Two competing
hypotheses were set up to test the main effects and the interaction effect of emotional responses and coping on
encounter satisfaction:
H4a: Emotional responses and coping have significant and separate effects on encounter satisfaction.
Emotional responses are positively related to encounter satisfaction and coping is negatively related to
encounter satisfaction.
H4b: Emotional responses and coping interact with each other to affect encounter satisfaction such that the
effect of emotional responses on encounter satisfaction is more pronounced when problem-focused coping
strategies are dominant (as compared to emotion-focused coping strategies).
The conceptual representation of the two hypotheses is illustrated in Figure 2.
Figure 2
The Conceptual Representation of the Hypothesized Relationships
Method
Subjects, Design, Procedure, and Experimental Conditions
A total of 152 people participated in Study 2. Participants were recruited using a database provided by
the human resources office at a major university in the Northeastern region of the United States. Questionnaires
were sent to 1,200 people randomly selected from the database. The returned questionnaires represent a
response rate of 12.5 percent. Of these, ninety four participants were female (62 percent). The participants
were between the ages of 19 and 68 (M = 45.8).
As coping strategies are most relevant in stressful service encounters, Study 2 was limited to negative
behaviors of other customers. Participants were exposed to a hypothetical consumption episode similar to the
one used in Study 1.
Dependent Variables
Emotional responses were the same measure used in Study 1.
Coping. Nine coping items relevant to this research were chosen from the COPE inventory (Carver, Scheier, &
Weintraub, 1989). The scale included six items representing problem-focused coping (e.g, “I would take action
to try to get rid of the problem” and “I concentrate my efforts on doing something about it”) and four items
reflecting emotion-focused coping (e.g., “I would look for something good in what is happening” and “I learn to
live with it”). The measurement was on a 7-point likert scale (1 = “I wouldn’t do this at all, and 7 = “I would do
this a lot”). The Cronbach’s Alpha of the six problem-focused coping items was 0.82 and the Cronbach’s Alpha
of the four emotion-focused coping items was 0.80. A coping index was formed by first averaging the problemfocused coping ratings and emotion-focused coping ratings separately and then subtracting the average emotionfocused coping from the average problem-focused coping, so that a greater score represents a greater tendency
to engage in problem-focused coping.
Encounter Satisfaction. Encounter satisfaction (α = .97) was measured using a six-item satisfaction scale
(Oliver & Swan, 1989). The satisfaction scale was anchored at 1 = very dissatisfied, and 7 = very satisfied.
Manipulation check variables and demographic variables were assessed with similar measurements
used in Study 1.
Results
Manipulation Checks. The scenario manipulations were first checked for their realism (M=5.42, SD=1.53) and
ease of use (M=5.77, SD=1.21).
Coping Strategies. Paired-sample t-test was performed to compare the salience of problem-focused coping
strategies and emotion-focused coping strategies participants used to deal with behaviors of others. The t-test
results showed that, as expected, individuals reported greater use of emotion-focused strategies than problemfocused strategies. was significantly stronger than appetitive goal (Memotion = 4.00, SD = 1.54 vs. Mproblem = 3.45,
SD = 1.42; t(147) = 2.23, p < .05). Therefore, Hypothesis 3 was supported.
The Relationships among Emotional Responses, Coping and Encounter Satisfaction.
To examine the effects of emotional responses and coping on encounter satisfaction, two multiple
linear regression models were tested. Model 1 regressed encounter satisfaction on emotional responses and
coping. Model 2 added an interaction term between emotional responses and coping to the model. The two
predictor variables were standardized into Z-scores and an interaction variable was created by multiplying the
two standardized variables (Bauer & Curran, 2005). The multiple regression modeling results showed that both
models showed significant goodness of fit (R2 = .41; F(2,137) = 47.94, p < .001 for Model 1; R2 =.45, F(3,136)
= 26.36, p < .001 for Model 2). The results further showed that the increment in the total variance explained by
Model 2 over Model 1 was statistically significant (∆ R2 = .04, F (1, 136) = 8,17, p < .01), suggesting a better
model fit of Model 2 that predicts a moderating effect of coping strategies on the effect of emotional responses
on encounter satisfaction. The regression results suggested that emotional responses, coping, and the interaction
between the two variables altogether explained 45 percent of the variance in encounter satisfaction (F(3,136) =
26.36, p < .001). The regression results of the both models are presented in Table 2.
To further probe the significant interaction effect in Model 2, we followed the J-N technique of testing
the simple slopes (Bauer & Curran, 2005). Specifically, we computed the simple intercepts and simple slopes
for the regression of encounter satisfaction on coping strategies at one standard deviation above and below the
mean, that is, z-scores of -1 (representing a dominant emotion-focused coping strategy) and 1 (representing a
dominant problem-focused coping strategy) were entered in the model to test the two simple slopes for emotionfocused and problem-focused copying strategies respectively, producing the plots in Figure 3. Specifically:
1) Equation 1: the general equation for Model 2:
Sat = 2.92 + .51Emotion + (-.39) Coping + .16 Emotion x Coping
2) Equation 2: the equation for emotion-focused coping strategy:
Sat = 2.92 + .51Emotion + (-.39) x (-1) + .16 Emotion x (-1)
= 3.31 + .35Emotion
3) Equation 3: the equation for problem-focused coping strategy:
Sat = 2.92 + .51Emotion + (-.39) x 1 + .16 Emotion x 1
=2.53 + .67Emotion
As shown in Equation 2 & 3, the effect of emotional responses on encounter satisfaction was much
stronger when problem-focused coping strategies (b = .67) were dominant as compared to emotion-focused
coping strategies (b = .35). Taken together, Hypothesis 4 was supported.
Discussion
The results show that while consumers use both emotion-focused and problem-focused coping
strategies to deal with the behavior of other customers, emotion-focused coping is more predominant than
problem-focused coping. The results also support the proposition that other-consumer-elicited moment-tomoment emotional responses and coping are integrated into encounter satisfaction. The two consumer
responses and the interaction between the two explain a considerable portion of the variance in encounter
satisfaction. Another notable finding from the regression analyses is that problem-focused coping accentuates
the effect of emotional responses on encounter satisfaction. Coping literature suggests that problem-focused
coping is generally more effective than emotion-focused coping in stressful encounters (e.g., Carver, Scheier &
Weintraub, 1989; Folkman et al., 1986; Lazarus & Folkman, 1984). Interestingly, this study finds that in coping
with other-consumer-elicited negative emotions, as reflected in the reduced level of encounter satisfaction,
problem-focused coping appears to be more taxing than emotion-focused coping. This finding is another piece
of empirical evidence of the psychological costs associated with encounters with other consumers in service
settings. Problem-focused coping, as an “oppositional approach” to stressful service encounters (Ringberg,
Odekerken-Schroder & Christensen, 2007), compromises encounter satisfaction. These results highlight the
importance of isolating other-consumer-elicited emotions from other emotion-eliciting sources in a service
environment (e.g., atmospheric factors and service providers) and of examining the unique underlying
psychological processes associated with other-consumer-elicited emotions.
Table 2
Multiple Linear Regression Results
Theoretical Implications
This research contributes to an emerging stream of consumer research that investigates the influence of
other consumers in service encounters. One important contribution of the current study to the literature is that
the findings capture the social nature of consumer-to-consumer interactions in service encounters. Findings
from this study show that emotional responses to even the desirable behavior of other customers are less positive
than the responses to behaviors that are typical and as expected. Furthermore, consumers inhibit emotional
displays of even positive feelings and often engage in passive emotion-focused coping to deal with the negative
behavior of others. The empirical evidence of the social inhibition effect of the presence of other customers on
consumer behavior revealed in this study sheds light on the psychological implications of the presence of other
consumers in service encounters. This attempt on psychological implications extends the consumer-toconsumer (C2C) research beyond its traditional focus on other consumers’ physical and behavioral impact on
consumption experiences.
Figure 3
The Regression Slopes by Coping Strategy
Implications and Future Research
The findings from this study suggest that the behavioral patterns of customers in response to the
behavior of others at service encounters tend to be restrained and defensive, as reflected in inhibited emotional
display and use of emotion-focused coping strategies. Such defensive approach turns out to be emotionally
taxing and costly, as revealed in the reduced level of encounter satisfaction attributed to the psychological costs
associated with expression inhibition and coping. The social inhibition effect emerged in this study further
demonstrates that the human triad in service encounters (a focal consumer-service provider-other consumers)
weaves a complex psychological landscape in which consumptions take place. It points to the importance of
examining the dyadic relationships in the human triad separately (e.g., consumer-to-consumer interaction and
consumer-to-provider interaction), as the nature of the two dyadic relationships is different (McCallum &
Harrison, 1985; Price & Arnould, 1999). On the other hand, it also highlights the need to integrate our
understanding of the dyadic relationships and to take a holistic perspective of human interactions in service
encounters.
Managerial Implications
The current research offers several important implications for hospitality practitioners. One clear
message for hospitality mangers is that in managing customer-to-customer interaction experience at service
encounters, clear behavioral rules need to be effectively communicated to all customers. In particular, unlike
customer-provider interactions, it is more important to encourage “normal” behaviors than “positive” behaviors
in customer-to-customer interactions. An encounter in which other customers behave as expected generates the
most favorable emotional responses.
Of particular pertinence to hospitality practitioners, this study shed light on consumer coping in
stressful service encounters. The results imply that consumers do expect service providers to intervene in
response to other consumers’ undesirable behaviors. If no effective actions are taken by service providers,
consumers may need to initiate problem-focused strategies to deal with other consumers’ negative behaviors.
However, the need to act on the part of the consumers will negatively affect encounter satisfaction and reduce
the overall evaluation of the consumption experience.
Limitations and Future Research
This study is not without limitations. First, this study may have traded away some internal validity in
order to use scenarios as stimulus material. Future research using field studies can validate the results of this
study. Second, the limitations of self-reported emotional response measures should also be acknowledged.
Although earlier studies demonstrate that self-reports provide an effective and efficient method of assessment in
measuring consumption-related emotions (Westbrook & Oliver, 1991), it would have been preferable to use
some observational measures such as facial expressions and actual actions taken to assess some consumer
responses, such as emotive dissonance and coping strategies. Third, this study was limited in the range of
services employed. Only restaurant services were empirically tested in this study. Using a wider range of
services whose usage is more tied to particular cultural and demographic groupings might have proved fruitful.
Replication of the results observed in this study using a wider range of services would also increase the
generalizability of the study findings.
References (Selected references are listed due to the page limit. A full reference list is available upon request.)
Bauer, D. J., & Curran, P. J. (2005). Probing interactions in fixed and multilevel regression: Inferenetial and
graphical techniques. Multivariate Behavioral Research, 40, 373-400.
Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies: A theoretically based
approach. Journal of Personality and Social Psychology, 56(2), 267-283.
Grove, S. J., & Fisk, R. P. (1997). The impact of other customers on service experiences: A critical incident
examination of "getting along". Journal of Retailing, 73(1), 63-85.