Journal of Retailing and Consumer Services 19 (2012) 211–217
Contents lists available at SciVerse ScienceDirect
Journal of Retailing and Consumer Services journal homepage: www.elsevier.com/locate/jretconser
Moderating effect of discriminatory attributions on repatronage intentions Thomas L. Baker a, Tracy Meyer b,n a b
College of Business and Behavioural Sciences, Clemson University, 249 Sirrine Hall, Box 341325, Clemson, SC 29634-1325, USA Cameron School of Business, University of North Carolina Wilmington, 601 S. College Road, Wilmington, NC 28403-5969, USA
a r t i c l e i n f o
abstract
Available online 7 February 2012
This research investigates the role of social context at the time of a service failure. The negativity of the service failure has the potential to escalate when consumers who are part of a traditionally stigmatized group believe the service failure to be a purposeful event brought on by physically observable differences in appearance. A small exploratory study and two large scale data collections are conducted to validate a measure of transaction specific attributions of discrimination (TSAD) to enhance our understanding of the issue. The moderating effect of TSAD is demonstrated on repatronage intentions to the service firm. Published by Elsevier Ltd.
Keywords: Service failure Repatronage behaviors Discrimination
1. Introduction The negativity surrounding a service failure has the potential to escalate when the customer believes that a difference between the parties (social context) played a role (Baker et al., 2008). These attributions of discrimination have serious implications for firm performance that can be best illustrated with examples. In 2001 twenty-one Black people filed a $100 million federal lawsuit against Cracker Barrel restaurants alleging they were denied service and segregated in the smoking section. Three Black men in Kips Bay, New York, filed a lawsuit against a McDonald’s restaurant in 2010 for being denied access to a customer only restroom when other (non-Black) individuals were allowed access. In both cases, perceptions of biased treatment based on conclusions drawn from contextual factors led to negative publicity and financial losses for the firm. Attribution theory suggests that one judges an outcome on what is believed to have been the cause of an incident and this information is then used to engender future actions (Weiner, 2000). Therefore, the ultimate effect of a failure will vary based on the attribution that arises (Iglesias, 2009). It is possible that the consumer may attribute the failure to a purposeful act and not merely a random incident. In the attribution literature these purposeful failures are described as being high in locus of responsibility (a combination of locus of causality and a locus of controllability – see for example Folkes, 1984; Weiner, 2000). That is, the cause of the failure is believed to be the fault of the service employee, resides within the employee, and was controllable. If information available at the time of the failure is such that
n
Corresponding author. Tel.: þ1 910 962 7202; fax: þ1 910 962 7983. E-mail address:
[email protected] (T. Meyer).
0969-6989/$ - see front matter Published by Elsevier Ltd. doi:10.1016/j.jretconser.2012.01.002
it causes the consumer to attribute the cause of the failure to the purposeful actions of a frontline employee then the failure is high in locus of responsibility. Consumers notice and react to the appearance of the frontline employee (Soderlund and Julander, 2009; Kim et al., 2009). We take this a step further and suggest that when the appearance of the employee and other customers present during the encounter is such that physical differences are evident, these differences have the potential to impact service perceptions. In the case of a service failure these readily observable physical differences have the ability to escalate the negativity of the situation. This is evidenced in anecdotal statements such as, ‘‘the service was slow because the server assumed that due to my young age I wouldn’t be a good tipper,’’ or ‘‘my meal was incorrectly prepared because I am Black and I’m not welcome here.’’ These discriminatory attributions are suspected to have an intense negative influence on the consumer. Therefore, consideration of these readily observable physical differences (social context) in a service failure warrants careful attention (Kulik and Holbrook, 2000). We suggest that the potential to attribute purposeful failures to observable individual differences reflects what has been labeled by Crockett et al. (2003) as marketplace discrimination. They define marketplace discrimination as ‘‘differential treatment of customers in the marketplace based on perceived group-level traits that produce outcomes favourable to ‘in-groups’ and unfavourable to ‘out-groups.’’’ Marketplace discrimination has been shown to occur over a number of different categories, including ethnicity, gender, age, and sexual orientation (Chung-Herrera et al., 2010; Walsh, 2009). Essentially, this research concludes that perceptions of marketplace discrimination act to create strong reactions to situations in which some observable difference is believed to be a factor. To date, most research in marketing focusing on discrimination and discriminatory attributions has been conducted relative
212
T.L. Baker, T. Meyer / Journal of Retailing and Consumer Services 19 (2012) 211–217
to racial or ethnic discrimination (see for example Rosenbaum and Montoya, 2007; Baker et al., 2008); however marketplace discrimination based on age, gender, disability, and sexual orientation also occurs. And while some have argued that the incidence of discrimination has declined over the past decades, Deitch et al. (2003) argue that overt discrimination has been replaced by more subtle yet pervasive acts. This covert form of discrimination may be more harmful due to the ‘‘attributional ambiguity’’ (Crocker and Major, 1989) associated with discrimination. Evidence of covert discrimination has been found in studies relating to obesity (King et al., 2006), religion (King and Ahmad, 2010), and ethnicity (Schreer et al., 2009) among others. One issue that might inhibit research in this area is the lack of measure designed to capture perceptions of marketplace discrimination that may be evident in a specific service encounter. The measure reported in this study that we term the ‘‘transaction specific attributions of discrimination’’ (TSAD) scale is specifically created to address this concern. We define TSAD as the attribution of an uncertain specific negative encounter to discrimination due to group membership. Most measures of discrimination that currently exist might best be considered measures of long-term discrimination in that they assess the extent to which a person has been exposed to discrimination over a period of time. These measures do not allow researchers to assess if a consumer’s perception of a specific transaction is being attributed to discrimination. While a comprehensive review of these measures is not possible, given space limitations, interested readers are directed to Utsey (1998). We believe there is potentially a broad application for a measure like the one developed here, particularly when the purpose is to study specific interactions between customers and employees. For example, within the services area any study that investigates perceptions of service failures and/or service recovery might incorporate discriminatory attributions as a potential causal variable for the perception that a failure has occurred. This is also true in sales, whether it be retail or business to business where interactions between individuals can easily be misconstrued. The primary objective of this paper is to examine the moderating role of discriminatory attributions on the relationship between trust and interactional justice and repurchase intentions. Trust and interactional justice were chosen due to the important roles they play in a service setting. However, this research adds to our knowledge of these important relationships by including the impact of discriminatory attributions. By doing so service firms should be better positioned to understand consumer’s perceptions of events surrounding service failures that include individual differences between the parties involved. This, in turn, should aid in the firm’s ability to engage in appropriate recovery tactics. The second objective is to create and validate a measure of transaction specific attributions of discrimination (TSAD) that will make it easier for this construct to be included in other studies.
2. Theoretical framework and hypotheses 2.1. Interactional justice, trust, and repatronage intentions Two concepts that have received a great deal of attention relative to consumer reactions to service failures are interactional justice and trust. Tax et al. (1998) define interactional justice as ‘‘fairness of the interpersonal treatment people receive during the enactment of procedures’’ (p. 62). Interactional justice is one of the three dimensions of justice. The other two dimensions include distributive justice and procedural justice. According to Blodgett et al. (1997) distributive justice relates to the fairness of the
outcome while procedural justice reflects the fairness of the policies and procedures by which the outcome is produced. The choice of interactional justice as opposed to distributive or procedural justice stems from the timing. The purpose of this research is to understand perceptions that form as the encounter progresses: specifically perceptions of discriminatory treatment that occurs during the experience. Interactional justice is positively related to a number of different constructs, including satisfaction with service provision (Clark et al., 2009; Martinez-Tur et al., 2006), satisfaction with complaint handling (Tax et al., 1998), and a greater willingness for consumers to engage in the co-production of services (Seigyoung et al., 2007). More importantly in the current context, interactional justice has been found to be related to repatronage intentions (Blodgett et al., 1997). Trust is defined by Sirdeshmukh et al. (2002) as ‘‘the expectations held by the consumer that the service provider is dependable and can be relied on to deliver on its promises’’ (p. 17). Although trust is generally viewed in terms of an on-going relationship, it has also been shown to develop even after brief encounters based on an observation of events as they transpire (Swan and Nolan, 1985; Wood et al., 2008). Many studies related to trust consider the consequences of trust relative to firm loyalty (see for example Tax et al., 1998; Sirdeshmukh et al., 2002; Gupta and Kabadayi, 2010). In addition, trust has been shown to be related to repatronage in a number of contexts including retailing (Zboja and Voorhees, 2006) and services (Eisingerich and Bell, 2007). Repatronage intentions refer to the response by consumers following the dissatisfying service experience specifically related to the desire to patronize the firm in the future (Oliver and Swan, 1989). By understanding one’s intentions to return to the firm you are considering the likelihood that the person involved in the incident would or would not consider giving the firm another chance (Grace and Cass, 2005; Homburg and Giering, 2001). Consistent with this research we hypothesize: H1. Interactional justice will be positively related to repatronage intentions. H2. Trust will be positively related to repatronage intentions. 2.2. Moderating effect of discriminatory attributions According to attribution theory, consumers will come to some conclusion as to why an event occurred, particularly when the event is unexpected and/or negative (Heider, 1958; Harris et al., 2006). As it relates to service failures, the ambiguous nature of the reason for or cause of many service failures will lead the consumer to rely on any cue available at the time of the exchange to make sense of the situation and assess blame. This ‘‘attributional ambiguity’’ can be experienced by almost anyone but is more likely to occur among those who are members of chronically stigmatized groups (Major and Crocker, 1993). Thus, it is important to note that in addition to ethnicity, attributions of discrimination can be made by any group, including, but not limited to, those who might attribute negative events to group membership based age, gender, weight, religion, having a disability, or sexual orientation. While covert acts of discrimination are likely to occur, the consumer involved in the situation has no way to know for sure if the reason for the negative outcome was biased. Thus, any information available at the time of the encounter, no matter how minor, may be used to understand the reason for the outcome. To better understand what information might be used and how it may be used, Walsh (2009) conducted a retrospective
T.L. Baker, T. Meyer / Journal of Retailing and Consumer Services 19 (2012) 211–217
qualitative study across five groups of consumers who are likely to experience marketplace discrimination (ethnicity, gender, age, disability, and sexual orientation). Examples of behaviors reported as being perceived as discriminatory included strange or disparaging looks, an unwelcoming environment, and being ignored. However, as minor as these may seem, more than half of those interviewed indicated they spread negative word of mouth to vent their feelings and to prevent someone from having the same experience. Schreer et al. (2009) had Black participants try on sunglasses in front of a White employee. The Black respondents reported being the subject of subtle forms of interpersonal bias in the form of being treated suspiciously by the White employee (e.g., stared at) and being followed through the store. The influence of demographic differences on service failure and recovery efforts was studied by Chung-Herrera et al. (2010). Results revealed that when perceptions of age or ethnic discrimination were evident, customers were much more likely to attribute a failure to the employee as opposed to the firm and the failure was deemed personal in nature. From the foregoing discussion a number of conclusions can be drawn. First, when service failures occur the inherent ambiguity will cause consumers to look to any cues available to aid in determining why the failure occurred and physical differences constitute one such factor. Second, what has been labeled marketplace discrimination occurs and may be widespread. Third, when discriminatory attributions are made relative to the cause of a service failure, the result may be more severe than when other attributions are made. Based on this we conclude that any positive relationships that might exist as hypothesized in H1 and H2 will be mitigated to some extent when discriminatory attributions are made. Therefore, H3. TSAD will moderate the relationship between interactional justice and repatronage intentions such that as TSAD increases the relationship between interactional justice and repatronage intentions will be attenuated. H4. TSAD will moderate the relationship between trust and repatronage intentions such that as TSAD increases the relationship between trust and repatronage intentions will be attenuated. 3. Research design 3.1. TSAD scale development As mentioned previously, one objective of this study is to develop a scale designed to assess attributions of discrimination occurring within a specific encounter. This is due in part to the fact that current measures of discrimination assess the extent to which a person has been exposed to actions that might be attributed to discrimination over a period of time. An example of such a measure is the Experiences of Discrimination Scale developed by Murry et al. (2001), which assesses the extent to which a respondent has been exposed to discrimination by asking them to indicate the number of times each of thirteen specific events has happened to them. Other similar measures include the Perceived Racism Scale (McNeilly et al., 1996) and the Index of Race Related Stress (Utsey and Ponterotto, 1996). While these would no doubt prove useful as ways to assess the extent to which someone believes to have been exposed to discriminatory actions they do not allow the determination of how a specific negative event might be attributed to discrimination. The first task involved a qualitative exploratory study conducted with 30 students. They were first asked to define discrimination and to then describe a service/retail encounter where they felt discriminated against, and finally to detail why
213
they felt the employee treated them that way. Several themes were identified in the responses, including being treated differently based on physical differences, specifically being young. A second task involved a comprehensive review of the literature pertaining to discrimination. Based on the exploratory study and the literature review, a total of 32 potential items were identified. The items were evaluated by three people who have conducted discrimination research in different fields, including marketing, social psychology, and counseling. Reviewers were asked to evaluate the extent to which each item (1) represented the conceptual domain of the construct, (2) were not duplicating one another, and (3) could be utilized across different situations. This led to the deletion of 19 items. The 13 that remained were utilized in an empirical study designed to validate the measure. 3.2. Study 1 – scale development (age discrimination) The objective of the first study was to assess the psychometric properties of the 13 items developed previously (see Table 1). The sample was 409 students from a mid-sized Southeastern university who participated as part of their course credit. Since the sample was composed of students we chose to use age discrimination as the context for the data collection. We could have had utilized a critical incident methodology in which we would have asked the respondents to think about a time where they believed they had been discriminated against due to their age; however, we were not convinced all students would be able to do so nor were we sure they would use a context that was appropriate for our study. Therefore, we chose to utilize a scenario approach in order to ensure that all respondents were using the same frame of reference. The survey was created using an online survey tool and was distributed to the students via e-mail. Upon opening the survey they were first provided with a ‘‘welcome’’ page, which provided an overview of the study, asked for their participation, provided instructions, and addressed issues related to respondent anonymity. They were then presented with the scenario (see Appendix A), which was adapted from previous research (Baker et al., 2008) and described a situation in which the respondent and a date decided to go out for dinner. They were explicitly requested to imagine themselves in that situation. The scenario details a significant delay in the food being served (service failure). Slow service was chosen as the failure due to the commonly experienced nature of the failure (Helms and Mayo, 2008; Kelley et al., 1993). In addition, two photos (see Appendix A) were provided in order to manipulate the issue of age. One photo portrayed the ‘‘waiter’’ in the scenario and the other was of a group of older customers having dinner. The pictures were pretested to be sure the manipulation worked such that student participants observed that the parties were indeed older. Bateson and Hui (1992) found that photos were able to invoke the same psychological feelings as an actual service setting. Furthermore, this methodology has been used successfully in other service research (Baker et al., 2008). After reviewing the scenario and pictures, they were asked to complete the items designed to assess TSAD using a 7 point Likert scale (1 ¼ strongly disagree; 7 ¼strongly agree). We also included items designed to assess the extent to which the scenario was believable and realistic. Those two items were measured using a 5-point Likert scale (1¼strongly disagree; 5¼strongly agree). The mean for believability item was 4.00 and for the realistic item 3.92 both of which provide evidence that the scenario was viewed as being believable and realistic. The assessment of the scale was conducted via a confirmatory factor analysis (CFA) using LISREL (Joreskog and Sorbom, 2001). Following suggestions by Hu and Bentler (1999) we augmented
214
T.L. Baker, T. Meyer / Journal of Retailing and Consumer Services 19 (2012) 211–217
Table 1 Summary of scale validation. Item
Factor loadingsa
l valuesb (tvalues) from Study 1
l valuesb (tvalues) from Study 2
1. 2. 3. 4. 5. 6. 7. 8.
.85 .71 .89 .78 .89 .68 .92 .91
.84c (12.96) .68 (10.00) .90 (13.63) .75 (11.88) .88 (13.47) .63 (d) .92 (13.890) .90 (13.69)
.84 .87 .92 .95 .93
.87 .90 .82 .85 .86
.85 (13.11)
e
If I were older (White), this failure would not have occurred I believe the service failure was a purposeful act of vengeance due to my age (race) The service failure occurred due to the server’s prejudice against my age (race) This failure is an indirect (covert) example of discrimination I was singled out for poor treatment because I am young (Black) The age (race) of the parties involved in the failure is relevant I received poor service because I am young (Black) I was discriminated against because I am young (because of my membership in the African-American race) 9. I was purposely treated poorly in this situation because of my age (race) 10. The poor service was due to prejudice based on age (race) 11. The poor service is another act of discriminatory behaviors 12. This service failure was due to age (racial) discrimination 13. This subtle slight is an act of discrimination Average variance extracted Composite reliability
(d) (19.23) (21.05) (22.60) (21.62)
e e e
f f f f
.63 .97
.81 .98
a
Factor loadings from exploratory factor analysis. Values from confirmatory factor analysis. c Completely standardized loadings. d No t-value since item was constrained to 1 to set the scale metric. e Items deleted due to confirmatory factor analysis in Study 2. f Items deleted due to confirmatory factor analysis in Study 1. b
the w2 statistic with one relative fit measure (RMSEA) and two absolute fit measures (CFI and IFI). In addition, as is convention we report average variance extracted and composite reliability for our scale. The initial model fit provided what could be considered limited support for the scale. The w2 statistic was 356.22(df ¼ 65, p o .05). RMSEA was .12, which is above the level that would provide confidence of good fit. However, both CFI (.94) and IFI (.94) were better than the suggested cut-off value of .90 (Hu and Bentler, 1999). To further evaluate the model we looked at a number of additional diagnostics to try to identify poorly performing items. We were careful to take into account the nature of specific items to ensure that the entire domain of TSAD was being measured. In addition to needing to find a model that fit, we wanted to identify as parsimonious a set of items as possible. Items were removed sequentially and the model re-estimated until arriving at a model that met the aforementioned goals. This resulted in a 5 item measure. The w2 statistic was 6.85(df ¼ 5, p ¼ .23), which is indicative of very good fit as are the CFI and IFI values of .99 and the RMSEA of .04. The completely standardized factor loadings and associated t-values are provided in Table 1. The average variance extracted value for the five items remaining was .68, which is above the recommended level of .50 and the composite reliability was very high at .98. 3.3. Study 2 – scale validation (racial discrimination) To provide further validation of the TSAD measure we changed the context in Study 2 to racial discrimination. The 149 respondents came from a firm that manages a national consumer panel. The same scenario used in Study 1 was used in Study 2 with the exception that ethnicity, rather than age, was manipulated via pictures of a White server and White customers (see Appendix A). As with the student sample, believability (5.54) and realism (5.64) scores were high based on a scale ranging from 1 (strongly disagree) to 7 (strongly agree). The five TSAD items that were identified in Study 1 were submitted to a confirmatory factor analysis. The model fit the data very well. The w2 statistic was 5.98 (df ¼ 5, p ¼ .31). The RMSEA value was .03 which is well within the range generally considered
as acceptable as were the values for CFI (.99) and IFI (.99). The factor loadings and t-values are provided in Table 1. The average variance extracted for the scale was .74 and the composite reliability was .96 both of which are above the suggested values of .50 and .70, respectively. Based on this replication of the initial analysis presented previously the items appear to have a high degree of validity and reliability. 3.4. Hypotheses tests In addition to items designed to allow a validation of the TSAD scale, Study 2 included items designed to test the four hypotheses between interactional justice, trust, and repatronage intentions. Interactional justice was measured using four items adapted from Tax et al. (1998). The four items were, ‘‘In dealing with my problem, the server treated me in a courteous manner,’’ ‘‘During their effort to fix my problem, the server showed a real interest in trying to be fair,’’ ‘‘The service employee worked as hard as possible for me during the recovery effort,’’ and ‘‘The server was honest and ethical in dealing with me during their fixing of my problem.’’ Responses were provided on a 7-point scale anchored by 1 (strongly disagree) and 7 (strongly agree). The four items which measured trust came from Sirdeshmukh et al. (2002). Response alternatives corresponded to the element of trust each item was designed to assess. Thus, the alternatives ranged from (1) ‘‘very undependable’’ to ‘‘very dependable,’’ (2) ‘‘very incompetent’’ to ‘‘very competent,’’ (3) ‘‘of very low integrity’’ to ‘‘of very high integrity,’’ and (4) ‘‘very unresponsive to customers’’ to ‘‘very responsive to customers.’’ A measure of repatronage intentions was adapted from Homburg and Giering (2001). Using a 7-point scale (1¼very low, 7¼very high), respondents were asked to indicate the probability of returning and the likelihood of making the same choice after the fact. The means, standard deviations, and coefficient alphas for each scale as well the correlation among the constructs are provided in Table 2. 4. Results To test the four hypotheses developed earlier, moderated regression analysis was conducted. As suggested by Aiken and West (1991)
T.L. Baker, T. Meyer / Journal of Retailing and Consumer Services 19 (2012) 211–217
215
Table 2 Means, standard deviations, coefficient alphas, and correlations for constructs in Study 2. Constructs
Mean
Std. deviation
Alpha
Repatronage
Trust
Interactional justice
TSAD
Repatronage Trust Interactional justice TSAD
3.30 4.50 5.28 2.29
1.51 1.33 1.23 1.54
.93 .93 .92 .96
1.00 .60a .51 .29
1.00 .67 .31
1.00 .28
1.00
a
All correlations significant at p o.001.
Repatronage Intentions
4.5
4
3.5 Low TSAD 3
High TSAD
2.5
2 Low Interactional Justice
High Interactional Justice
Fig. 1. Interaction plots. Plot for interactional justice. Fig. 2. Interaction plots. Plot for repatronage intentions.
the independent variable and moderator were mean-centered before the interaction term was created in order to reduce multicollinearity. Given our hypotheses, we expect there to be a positive coefficient for the independent variable (interactional justice, trust) that provides a test of H1 and H2 and a negative coefficient for the interaction term that provides a test of H3 and H4. Although not hypothesized, we also expect there to be a negative coefficient for TSAD that reflects the negative impact of TSAD on repatronage intentions. To test H1 and H3, we regressed repatronage intentions on interactional justice, TSAD, and the interaction between interactional justice and TSAD. The coefficient between interactional justice and repatronage intentions was .58 (po.00), which provides support for H1, and the coefficient between the interaction term and repatronage intentions was .18 (p ¼.01), which provides support for H3. Furthermore, the negative sign for the interaction term indicates that as TSAD is higher the relationship between interactional justice and repatronage intentions decreases. In other words when consumers attribute a service failure to attributions of discrimination the relationship between interactional justice and repatronage intentions is weaker. Further evidence for this relationship can be found in Fig. 1, which plots the relationship. Also, it should be noted that the coefficient for TSAD is .18 (p ¼.01), indicating that repatronage intentions decline as perceptions of TSAD increase. The tests of H2 and H4 were similar with the exception being that trust was included in the equation rather than interactional justice. The coefficient between trust and repatronage intentions was .63 (p o .00), providing support for H2, and the coefficient between the interaction term and repatronage intentions was .14 (p ¼.00), which supports H4. Again, this indicates that as TSAD gets higher the relationship between trust and repatronage intentions is lessened. Thus, TSAD moderates the relationship such that the relationship between trust and repatronage intentions is less when TSAD is high, which is what would be expected. This relationship is plotted in Fig. 2. Also, as with interactional justice, the coefficient for TSAD is negative ( .16, p ¼.02), which indicates that as TSAD increases, repatronage intentions decrease.
5. Discussion Discrimination, regardless of its basis, has been and will continue to be a critical issue in society. While most people involved in marketing would argue their firm or employees do not act in a discriminatory manner, that really is not the issue. The issue is the extent to which those who consider themselves to be part of a traditionally stigmatized group might attribute a service failure to discrimination in situations when the reason for the failure is ambiguous. Baker et al. (2008) have shown that this occurs and that it has real economic costs to the firm due in part to the higher levels of service recovery required after the failure. This research sheds additional light on the impact of discrimination by investigating how attributions of discrimination after a service failure moderate relationships between interactional justice and repatronage intentions and trust and repatronage intentions. In both cases, discriminatory attributions attenuated the aforementioned relationships, which are considered to be critical for the success of a service firm. Thus, it would appear that a better understanding of service encounters can be had by including perceptions of discrimination. This is particularly true with services marketing, which comes down to the interaction between the customer and the employee. In fact, it has been stated that for many consumers the interaction with a service employee is the service (Bitner et al., 1990). Accordingly, firms must be aware of how consumers are likely to react to negative service encounters that involve any ambiguity as to the cause particularly when there is the possibility of observable differences between the customer and the other parties present at the time of the encounter. Being blind to the feelings of others is not acceptable. Firms need to wake up to the fact that demographic differences can lead to more extreme negative outcomes. Simply understanding the negativity that can be derived from attributions related to physical differences is the first step. Another thing firms should consider is hiring a diverse group of employees, employees who physically reflect the customer base such that all customers will feel comfortable and social context will less likely be an issue.
216
T.L. Baker, T. Meyer / Journal of Retailing and Consumer Services 19 (2012) 211–217
Photo 1. Photos for age discrimination (Study 1).
Photo 2. Photos for ethnicity discrimination (Study 2).
In addition to implications for mangers, we believe that the TSAD measure has the potential to provide explanatory power to a wide range of service failure studies. Our conceptualization and operationalization of the measure at the transaction specific level is important given that most measures of discrimination assess what might better be labeled ‘‘long term’’ or ‘‘chronic’’ discrimination. Researchers will now have a multi-item, psychometrically sound measure that can be used to assess discriminatory attributions arising from a particular situation.
6. Limitations and future research While we feel comfortable that our development and assessment procedures were adequate, more work remains to be done. First, in both cases we collected the research using similar experimental manipulations. Using the scale in a non-experimental situation (e.g., a critical incident methodology) would be a fruitful next step. Second, our use of interactional justice, repatronage intentions, and trust appeared to be useful constructs for a test of nomological validity but the ability to test the final version of the scale relative to other constructs (e.g., behavioral outcomes) would provide more evidence of the efficacy of the scale. A limitation of this research involved the focus on discriminatory attributions from a US point of view. Broadening the scope of the research to an international perspective would add merit. Also experiments in which groups who receive different treatment are compared would provide additional insights. Finally, simply using the scale in studies that focus on different forms of potential discriminatory attributions will provide evidence of the extent to which the scale is as generalizable as we hope it to be.
Appendix A Scenario and photos You and three friends arrive at a casual dining restaurant for dinner (see photo below). The restaurant is crowded but you are able to find a table. You place your drink and food order with the server (see photo below). It takes the server nearly 10 min to bring you your drinks. After 40 min you still do not have your food but you notice that people who arrived after you are nearly finished with their meals. You call the server over to complain about the slow service. The server listens carefully to your complaint and is very respectful of what you have to say. The server then sincerely apologizes for the delay. In a few minutes the server returns with your food and refills your drinks. The server apologizes again for the slow service and states that when she put in your food order the ticket was misplaced in the kitchen. Once the kitchen personnel noticed the ticket they immediately prepared the food. You and your friends eat, pay the bill, and then leave the restaurant (Photos 1 and 2). References Aiken, L.S., West, S.G., 1991. Multiple Regression: Testing and Interpreting Interactions. Sage Publications, Newbury Park, CA. Baker, T.L., Meyer, T., Johnson, J.D., 2008. Individual differences in perceptions of service failure and recovery: the role of race and discriminatory bias. Journal of the Academy of Marketing Science 36, 552–564. Bateson, J.E.G., Hui, M.K., 1992. The ecological validity of photographic slides and videotapes in simulating the service setting. Journal of Consumer Research 19 (2), 271–281. Bitner, M.J., Booms, B.H., Tetreault, M.S., 1990. The service encounter: diagnosing favorable and unfavorable incidents. Journal of Marketing 54, 71–84.
T.L. Baker, T. Meyer / Journal of Retailing and Consumer Services 19 (2012) 211–217
Blodgett, J.G., Hill, D.J., Tax, S.S., 1997. The effects of distributive, procedural, and interactional justice on postcomplaint behavior. Journal of Retailing 73, 185–210. Chung-Herrera, B.G., Gonzalex, G.R., Hoffman, K.D., 2010. When demographic differences exist: an analysis of service failure and recovery among diverse participants. Journal of Services Marketing 24 (2), 128–141. Clark, M.N., Adjei, M.T., Yancy, D.N., 2009. The impact of service fairness perceptions on relationship quality. Journal of Services Marketing 30, 287–302. Crocker, J., Major, B., 1989. Social stigma and self-esteem: the self-protective properties of stigma. Psychological Review 96 (4), 608–630. Crockett, D., Grier, S.A., Williams, J.A., 2003. Coping with marketplace discrimination: an exploration of the experience of black men. Academy of Marketing Science Review, 7. Deitch, E.A., Barsky, A., Butz, R.M., Chan, S., Brief, A.P., Bradley, J.C., 2003. Subtle yet significant: the existence and impact of everyday racial discrimination in the workplace. Human Relations 56 (11), 1299–1324. Eisingerich, A.B., Bell, S.J., 2007. Maintaining customer relationships in high credence services. Journal of Services Marketing 21, 253–262. Folkes, V.S., 1984. Consumer reactions to product failure: an attributional approach. Journal of Consumer Research 10, 398–409. Grace, D., Cass, A.O., 2005. An examination of the antecedents of repatronage intentions across different retail store formats. Journal of Retailing and Consumer Services 12 (4), 227–243. Gupta, R., Kabadayi, S., 2010. The relationship between trusting beliefs and web site loyalty: the moderating role of consumer motives and flow. Psychology & Marketing 27, 166–185. Harris, K.E., Mohr, L.A., Bernhardt, K.L., 2006. Online service failures, consumer attributions and expectations. Journal of Services Marketing 20 (6/7), 453–458. Heider, F., 1958. The Psychology of Interpersonal Relations. Lawrence Erlbaum Associates, Hillsdale, NJ. Helms, M.M., Mayo, D.T., 2008. Assessing poor quality service perceptions of customer service representatives. Managing Service Quality 18 (6), 610–622. Homburg, C., Giering, A., 2001. Personal characteristics as moderators of the relationship between customer satisfaction and loyalty: an empirical analysis. Psychology and Marketing 18, 43–66. Hu, L., Bentler, P.M., 1999. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Structural Equation Modeling 6, 1–55. Iglesias, V., 2009. The attribution of service failures: effects on consumer satisfaction. The Service Industries Journal 29 (2), 127–141. Joreskog, K.G., Sorbom, D., 2001. Lisrel 8: User’s Reference Guide. Scientific Software International, Chicago, IL. Kelley, S.W., Hoffman, K.D., Davis, M.A., 1993. A typology of retail failures and recoveries. Journal of Retailing 69 (4), 429–452. Kim, J.-E., Ju, H.W., Johnson, K.K.P., 2009. Sales associate’s appearance: links to the consumers’ emotions, store image, and purchases. Journal of Retailing and Consumer Services 16, 407–413. King, E.B., Ahmad, A.S., 2010. An experimental field study of interpersonal discrimination toward Muslim job applicants. Personnel Psychology 63, 881–906. King, E.B., Shapiro, J.R., Hebl, M.R., Singletary, S.L., Turner, S., 2006. The stigma of obesity in customer service: a mechanism for remediation and bottom-line consequences of interpersonal discrimination. Journal of Applied Psychology 91 (3), 579–593.
217
Kulik, C.T., Holbrook Jr., R.L., 2000. Demographics in service encounters: effects of racial and gender congruence on perceived fairness. Social Justice Research 13 (4), 375–402. Major, B., Crocker, J., 1993. Social stigma: the affective consequences of attributional ambiguity. In: Mackie, D., Hamilton, D. (Eds.), Affect, Cognition, and Stereotyping: Interactive Processes in Group Perception. Elsevier, New York, pp. 345–370. Martinez-Tur, V., Peiro, J.M., Ramos, J., Moliner, C., 2006. Justice perceptions as predictors of customer satisfaction: the impact of distributive, procedural, and interactional justice. Journal of Applied Social Psychology 36, 100–119. McNeilly, M.D., Anderson, N.B., Armstead, C.A., Clark, R., Corbett, M., Robinson, E.L., Peiper, C.F., Lepisto, E.M., 1996. The perceived racism scale: a multidimensional assessment of the experience of white racism among African Americans. Ethnicity and Disease 6, 154–166. Murry, V.M., Brown, P.A., Brody, G.H., Cutrona, C.E., Simons, R.L., 2001. Racial discrimination as a moderator of the links among stress, maternal psychological functioning, and family relationships. Journal of Marriage and the Family 63, 915–926. Oliver, R.L., Swan, J.E., 1989. Consumer perceptions of interpersonal equity and satisfaction in transactions: a field survey approach. Journal of Marketing 53, 21–35. Rosenbaum, M.S., Montoya, D.Y., 2007. ‘‘Am I welcome here?’’ Exploring how ethnic consumers assess their place identity. Journal of Business Research 60 (3), 206–214. Schreer, G.E., et al., 2009. ‘‘Shopping while Black’’: examining racial discrimination in a retail setting. Journal of Applied Social Psychology 39 (6), 1432–1444. Seigyoung, A., Bell, S.J., McLeod, C.S., Shih, E., 2007. Co-production and customer loyalty in financial services. Journal of Retailing 83, 359–370. Sirdeshmukh, D., Singh, J., Sabol, B., 2002. Consumer trust, value, and loyalty in relational exchange. Journal of Marketing 66, 15–37. Soderlund, M., Julander, C.-R., 2009. Physical attractiveness of the service worker in the moment of truth and its effects on customer satisfaction. Journal of Retailing & Consumer Services 16 (3), 216–226. Swan, J.E., Nolan, J.J., 1985. Gaining customer trust: a conceptual guide for the salesperson. Journal of Personal Selling & Sales Management 5 (November), 39–48. Tax, S.T., Brown, S.W., Chandrashekaran, M., 1998. Customer evaluations of service complaint experiences: implications for relationship marketing. Journal of Marketing 62, 60–76. Utsey, S.O., 1998. Assessing the stressful effects of racism: a review of instrumentation. Journal of Black Psychology 24 (3), 269–288. Utsey, S.O., Ponterotto, J.G., 1996. Development and assessment of the index of race-related stress (IRRS). Journal of Counseling Psychology 43, 1–12. Walsh, G., 2009. Disadvantaged consumers’ experiences of marketplace discrimination in customer services. Journal of Marketing Management 25 (1–2), 143–169. Weiner, B., 2000. Attributional thoughts about consumer behavior. Journal of Consumer Research 27 (Dec), 382–387. Wood, J.A., Boles, J.S., Babin, B.J., 2008. The formation of buyer’s trust of the seller in an initial sales encounter. Journal of Marketing Theory and Practice 16 (Winter), 27–39. Zboja, J.J., Voorhees, C.M., 2006. The impact of brand trust and satisfaction on retailer repurchase intentions. Journal of Services Marketing 20, 381–390.