International Journal of Hospitality Management 35 (2013) 180–192

Contents lists available at ScienceDirect

International Journal of Hospitality Management

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

Effectiveness of recovery actions on deviant customer behavior—The

moderating role of gender
Huey Chern Boo a,∗ , Anna S. Mattila b , Ching Yeng Tan a
a
Faculty of Food Science and Technology, Universiti Putra Malaysia, Serdang 43400, Selangor, Malaysia
b
School of Hospitality Management, Pennsylvania State University, 224 Mateer Building, University Park, PA 16801, USA

a r t i c l e i n f o a b s t r a c t

Keywords: This paper aims to examine how gender interacts with various recovery actions on perceived employee
Deviant customer effort, complaint handling satisfaction, and behavioral intention in the case of deviant customer behav-
Deviant behavior ior in an East-Asian context. Study 1, a qualitative study, reveals that consumers expect companies to
Approach-avoidance
take appropriate remedial actions when a service consumption experience is affected by deviant cus-
Perceived effort
tomer behaviors (e.g., smoking in non designated area). Study 2 employed a 2 (gender) × 6(recovery
Service recovery
Service failure actions) between subject design in a restaurant setting. Results show that confronting the deviant cus-
tomer received the highest customer ratings in terms of perceived employee effort, complaint handling
satisfaction, and behavioral intention. However, gender moderated the relative effectiveness of recovery
actions. Relocating the focal customer was equally good as compensation for women while men were
less appreciative of relocation than compensation. In sum, service providers are advised to initiate some
form of recovery action in response to deviant customer behaviors.
© 2013 Elsevier Ltd. All rights reserved.

1. Introduction actions in addition to the more conventional recovery methods

(i.e., apology and compensation). Second, we propose that con-
Most services are delivered in the same location in which they sumers will react differently to service recovery efforts for deviant
are produced. Thus, the presence of fellow customers in the service customer behaviors depending on their gender. Over the years,
delivery is well-recognized in the service literature (Grove and consumer researchers have investigated a number of interesting
Fisk, 1997; Lovelock, 1994; Martin and Pranter, 1989). The adverse issues relating to gender (e.g., Schmitt et al., 1988; Fischer and
financial, physical, psychological, and social costs to organizations, Arnold, 1990). Yet, research focusing on the impact of gender on
service employees, and other consumers caused by deviant cus- customer evaluations of service encounters is sparse (Mattila et al.,
tomer behaviors have been widely documented (Fisk et al., 2010; 2003, 2009). We first conducted a qualitative study to better under-
Fullerton and Punj, 1993; Harris and Reynolds, 2003; Reynolds stand the types of deviant behaviors and customer expectations
and Harris, 2006). A deviant customer behavior can be defined as in an East-Asian context—Malaysia and then examined the gender
the conduct of customer violating the norms and unwritten rules differences in an experimental setting.
of a service setting in a negative fashion (Reynolds and Harris,
2009).
A review of the literature on deviant customer behaviors
revealed that research in a non-western context is scant (Huang, 2. Study 1
2008, 2010). In addition, to echo the call for further understanding
of culture in shaping customers’ reactions toward service failures 2.1. Background
(Kanousi, 2005; Wong, 2004), this research extended previous liter-
ature in two ways. First, we broadened the scope of recovery efforts In a service setting, customers might affect one another directly
by examining the effectiveness of approach-avoidance corrective (e.g., through specific interpersonal encounters as in support
groups) or indirectly (e.g., the enjoyment derived from crowds
at sporting events, in movie theatres, and in other entertainment
venues) (Baker, 1987; Bitner, 1992). Interactions with other cus-
∗ Corresponding author at: Department of Food Service and Management, Faculty
tomers can be positive or negative. For example, Arnould and Price
of Food Science and Technology, Universiti Putra Malaysia, Serdang 43400, Selangor,
Malaysia. Tel.: +603 8946 8364; fax: +603 8942 3552.
(1993) found that an interaction with others in a water rafting con-
E-mail addresses: hcboo@putra.upm.edu.my (H.C. Boo), asm6@psu.edu text had a profound positive effect upon one’s satisfaction. Unfortu-
(A.S. Mattila), valerie tcy@hotmail.com (C.Y. Tan). nately, customers frequently fail to fulfill their co-production roles.

0278-4319/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijhm.2013.06.005
 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192 181

Table 1 to indicate (a) if the complaint was handled satisfactorily and (b)
Typology of customer behavior.
the expected recovery action if the complaint handling was dissat-
Research Typology isfactory. If no complaint was voiced, participants were asked if the
Customer perspective situation could have been changed and if so what could have been
Lovelock (1994) Vandals (who intentionally deface organizational done.
property) A content analysis by two independent judges grouped the 150
Thieves (referring to customers who have no critical incidents (two of the 148 participants gave two incidents)
intention of paying for a service)
into the following categories: location, specific deviant behaviors,
Belligerents (represents customers who act in an
argumentative or aggressive fashion toward whether a complaint was made, satisfaction with complaint hand-
service personnel) ling, and expected recovery actions. Subsequently, a discussion was
Family feuders (who quarrel with other customers held between the two judges to revise the categorization scheme.
and family members)
The typologies documented in previous research (Table 1) were
Deadbeats (denotes customers who fail to pay for
services that they have already received) reviewed and both judges agreed on the following six groups of
Rule breakers (customers who fail to conform to deviant behaviors: (a) grungy (Martin, 1996), (b) inconsiderate
the unwritten rules and norms of service (Martin, 1996), (c) rule breaking (Bitner et al., 1994; Lovelock,
encounters) 1994), (d) crude (Martin, 1996), (e) violent or physical abuse (Bitner
Martin (1996) Gregarious (showing benevolent acts)
et al., 1994; Harris and Reynolds, 2004; Martin, 1996), and (f) ver-
Grungy (propping, smelly, and untidy)
Inconsiderate (showing disrespect for the rights or bal abuse (Bitner et al., 1994; Harris and Reynolds, 2004). In terms
feelings of others) of the actual and expected recovery actions, the judges formed
Crude (lacking taste, polish, or tact) six categories: no action, compensation, apology, relocation of
Violent (demonstrating excessive force or sudden
complainants, confronting the deviant customer, and others (pre-
intense behavior)
Malcontent (chronically dissatisfied)
ventive action by employees, companies, and government, promise
Leisurely (not overly time-conscious or rushed) of non-recurrence). The incidents were then reanalyzed and reas-
Grove and Fisk Protocol incidents reflecting violations of rules of signed into the agreed upon categories. By the end of the process,
(1997) conduct (physical incidents in line, verbal the inter-judge agreement for the categorization task was 0.84 and
incidents in line, other incidents in line, other
a full consensus was achieved for recovery actions. These values
protocol incidents)
Fellow customers’ sociability (friendly and exceed the 0.59 required proportion of inter-judge agreement to
unfriendly incidents, ambiance incidents) achieve the acceptable Nunnally’s reliability level of 0.70 (Rust and
Employee perspective Cooil, 1994). Disagreements on the deviant behavior categorization
Bitner et al. (1994) Drunkenness
were resolved through discussion.
Verbal and physical abuse
Breaking company policies
Below is an example to illustrate the categorization of deviant
Lack of cooperation behaviors and recovery actions. The incident took place in a depar-
Harris and Reynolds Compensation letter writers ture lounge (transportation). Two types of deviant behaviors were
(2004) identified, namely yelling (crude) and using abusive language (ver-
Undesirable customers
bal abuse). It is clear that the participant did not complain but
Property abusers
Service workers expected some action from the service staff to approach the deviant
Vindictive customers customer.
Oral abusers
Physical abusers My flight was delayed for more than 1 ½ hours. The [unfor-
Sexual predators tunate] incident happened when the passengers were waiting
to board the rescheduled flight.... a few passengers started to
form a line leading to the departure gate... and “a few lines”
More often than not, their behaviors contribute to other customers’ of waiting passengers also developed, creating a slightly messy
dissatisfaction (e.g., Bitner et al., 1994; Grove and Fisk, 1997). scene. All of the sudden, a yell was heard from behind the back
Over the years there have been many attempts to categorize of the pack of passengers. It was a very tall and big Caucasian. He
deviant customer behaviors (Bitner et al., 1994; Grove and Fisk, yelled at the passengers blocking his way, as he tried to squeeze
1997; Harris and Reynolds, 2004; Lovelock, 1994; Martin, 1996). through. He also used abusive language along the way... I was
These typologies are summarized in Table 1. shocked and disturbed by the man’s threatening [behaviors]. I
In sum, prior research shows that deviant behaviors reflect vary- believe both his strong physical build and use of abusive lan-
ing degrees of severity from passive complaining to active physical guage contributed to the threat I felt. Some other passengers
and sexual violence. In this study, we explore customers’ deviant were visibility disturbed. It was when the gate opened for board-
behaviors in Malaysia. To that end, Study 1 employed a critical ing, that I realized he had bought an “express boarding” pass. As
incident technique. the incident unfolded, no one from the airline’s ground crew
approached the man to find out the reason, if any, for his unac-
ceptable and threatening [behaviors]. I was disappointed with
2.2. Materials and methods
the airline staff’s serious lack of initiative to take control of the
situation.
Data collection was conducted via two modes: personal
interviews and self-administered open-ended questionnaires. To
produce a knowledgeable sample, a purposive sampling plan was
employed. Shoppers from a large mall in Klang Valley, a metropoli- 2.3. Results
tan area in Malaysia, were approached. Only those who indicated
that they had experienced a dissatisfactory service experience The majority of the participants were female (66%) and most of
caused by other customers were recruited. As a result, 148 par- the incidents occurred in a foodservice context (45%). Movie the-
ticipants were asked to elaborate on other customers’ deviant atres were the second largest group (n = 45) while the rest of the
behaviors. In addition, they were asked to indicate if a formal com- incidents included 13 in retail stores, six in banks, six in transporta-
plaint was lodged. If a complaint was filed, participants were asked tion companies (airlines and buses), two in hotels, and 10 in others
182 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192

Table 2
Consumer misbehaviors by category and service industry.

Misbehaviors Foodservices (n = 68) Hotels (n = 2) Transport (n = 6) Retails (n = 13) Theatres (n = 45) Banks (n = 6) Others (n = 10)

G1: Grungy
Unhygienic behavior 7 1
Taking off shoes 3
Propping feet 2
G2: Inconsiderate
Noise 14 1 7 1 1
Talking loudly 7 1 1 19 1 3
Late 1
Blocking ways 1 1
Unreasonable request 1
Ignore waiting customers 1 1 2
G3: Rule breaking
Using mobile phone 13 1 2
Breaking in line 7 10 2 3
Smoking cigarette 36 1 1 3 3
G4: Crude behavior
Drunk 5
Kissing 1
Quarrel 1
Yelling 1 1
G5: Violent
Fighting 1
Pushing or Kicking 3 12
Physically disturb others 2
G6: Verbal abuse
Foul language 3 1 2 1
Erotic words 1
Profanity 1 1

Note. The column total may be greater than the total number of incidents, in parenthesis, within each column since there could be more than one type of misbehavior in each
reported incident.

service establishments such as spas, scuba diving training centers, customers expect service providers to take action when the
and self-confident building workshops. deviant behavior clearly violates the stated rules or social norms
Among the types of deviant behaviors (Table 2), smoking was and is perceived to be controllable by the service provider. Cor-
the most disapproved behavior. Approximately 82% of the reported rective actions such as confronting the deviant customer seem to
smoking incidents took place in the foodservice setting (36 out result in satisfaction with problem handling. Other possible recov-
of 44). This behavior was condemned mainly because it occurred ery actions include apologizing, compensating, and moving the
in an air-conditioned or non-smoking area of a restaurant. Cus- focal customer to another area. Study 2 was conducted to address
tomers perceived that such a behavior is controllable by the service the following objectives: (1) to offer empirical evidence to the
provider as exemplified by these quotes: effectiveness of different recovery actions in a restaurant setting
and (2) to examine the role of gender in influencing consumers’
One customer sat [at the table] beside [us] was smoking . . ..
reactions to recovery efforts.
That restaurant is air-conditioned. . . . I found two signboards
Based on the results of Study 1, we are interested in the follow-
stating NO SMOKING. . . . I felt more [angry] because knowing
ing service recovery actions: compensation, apology, separation of
the customer [had broken] the rule, the service providers didn’t
the focal customer from the deviant customer, and correcting the
do their job well. They should stop the customer . . .
deviant behavior (i.e., smoking) exhibited by a fellow customer.
We also examined what kinds of action customers expect
service providers to take when faced with deviant customers. Only 3.1. Theoretical background
20 (13%) incidents were made known to the service provider.
Satisfaction with complaint handling was mixed (11 satisfactory Service recovery literature has documented that customer satis-
and 9 dissatisfactory). Satisfaction reflected action taken by the faction and future loyalty depend upon whether the customers felt
service staff or managers, such as appreciating the complaint, that they were treated fairly. Customer dissatisfaction is intensified
promising to improve, apologizing, and most importantly con- when (a) customer complaints are not given appropriate attention,
fronting the deviant customer. Conversely, dissatisfaction was (b) customers receive no apology, compensation, or explanation, or
mainly attributed to lack of action. (c) employees make no effort to resolve the problem (Bitner et al.,
Among the 130 non-complaint incidents, 69% of the respondents 1990). Furthermore, no recovery effort has been shown to lead to
anticipated some form of action from the service provider. Their low repurchase intention (Voorhees et al., 2006).
expectations reflected the need to uphold the rule or to set up the In the customer-to-customer interaction context, several
code of conduct, apologizing for the inconvenience caused, an offer researchers have suggested possible recovery actions. For instance,
to relocate the focal customer away from the deviant customer, and Huang (2010) revealed that customers’ post-recovery satisfac-
confronting the deviant customer. Table 3 presents some examples tion and subsequent behavioral intentions depend on perceived
of deviant behaviors and the actual or expected employee actions. employee effort and compensation. Jou and Yang (2007) also indi-
cated that a proper handling mechanism may require frontline
3. Study 2 employees or managers to stop the deviant behavior or to isolate
deviant customers from other functional customers, and to prop-
The findings from Study 1 indicate that smoking is the most erly compensate the affected customers. Huang (2008) denoted
common deviant behavior in a restaurant setting. Moreover, that the firm and its employees may express empathy, solve the
 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192 183

Table 3
Examples of deviant customer behaviors and employees’ actual and expected responses.

Incident Deviant behaviors Responses

It happened when I was dining at a Chinese restaurant . . .. the children were running and shouting in the Noise No action (A)
restaurant . . .. I [was irritated] and disappointed with the restaurant management. I [thought] they should apologize,
either politely ask[ed] the children [to] go back to their table or apologize[d] to us confrontation (E)
I went to a restaurant [with my family] . . . 2 or 3 months ago. We [chose to sit at] a table near to the aquarium. Noise, smoking Relocation (A)
Then [came] a group of Indians. [T]hey [were] just drinking beers, [making a lot of noise] and smoking . . ..
[We felt like leaving as soon as possible]. Luckily, the boss [relocated us] to a bigger table [further away]
When my friends and I dined in a Taiwanese Restaurant [t]he opposite table [was occupied] by a young Kissing No action (A)
couple. It was irritating [seeing the] couple kissing all the time. They didn’t realized or notice the reaction of confrontation (E)
other customers in the restaurant. We were ashamed [of them] and finished our meal very quickly. It was
sad [that not a] crew take[took] immediate action to stop [their] behavior
. . .. [It was on a buffet line]. The restaurant provided [one] scoop for each dish. But most customers [used] the Unhygienic behavior Others—preventive
same scoop to take other dishes . . .. [Not only it] contaminated [the dishes but also] mixed the meat [and] action,
non meat dishes. This [would cause] problem to the vegetarian [customers] . . .. I wished [the restaurant sign/guide/policy (E)
could have] provide[d] more scoops or spoons for each dish. [A] guideline or signboard can also be used to
teach the customers . . ..
. . . One of the customers [did] not switch off his mobile phone while others [were] enjoying their movie. Using mobile phone, Nothing (A, E)
Moreover, he raise[d] his voice volume and kick[ed] my seat . . . His act[ion] was just exactly [like] the [one talking loud, kicking
in] the Digi ad[vertisement] who [did not] switch off [his] phones in the movie hall. I talked to the customer
who kicked my seat. Only then he started to behave himself
I was queuing up to pay for the bread . . .. There were two or three customers in front of me and suddenly a Breaking in line No action (A)
Malay lady just cut the queue by walking directly to the cashier and [checking out]. I was irritated because Confrontation (E)
everyone was queuing up and waiting for their turns. [S]he [did not respect] everyone in the queue. I [did
not complain but] my [face] expression was [un]friendly. I purposely talked loudly to my mum and
mentioned the word “barbarian” so that she can hear. Cutting queue is very common and the cashier didn’t
say anything also. I expect the cashier ignore the lady or ask her to queue up
I rushed to [make a] cash withdrawal [during my lunch] break. [Both of the] ATM machine [were with] long Ignore waiting Others—Customer’s
queue. [W]hen [it] was the turn of the guy in front of me, I [thought it would be done in no] more than customers own awareness (E)
10 min. But [t]his guy [first checked the balance of his three cards, then made withdrawal from each of
them]. [T]he whole process [took] 25 min. I waited 35 min [in total]. . . . I didn’t complain because [it was his
right] though he may want to consider the people [lining] behind him
. . .. [My friend] and I did not have license [to dive], so we need to [attach to] an instructor. [We joined another Late Apology (A)
group of eight people]. [The lesson started] late [because the] 10 young members arrived [late without compensation (E)
feeling sorry for their action]. The staff apologized but did [not] compensate [us] . . .. I felt being cheated.
[T]hey could have at least offered to return [part of the] money or compensate [us] by [replacing another
day of lesson]

Note. (A) indicates the actual recovery action; (E) indicates the expected recovery action.

problem in an expedient manner, offer a heartfelt apology, or dis- 3.1.1.3. Problem solving—Corrective action or removal of the focal cus-
count. In sum, compensation, apology, separation of deviant and tomer. Although apology and compensation are effective means for
functional customers, and resolving misbehavior of other customer recovering from service failures (e.g., Grewal et al., 2008; Wirtz and
offer various remedies to deal with complaints arising from deviant Mattila, 2004), the root cause must be corrected. As pointed out by
customer behaviors. This array of recovery actions is consistent McColl-Kennedy and Sparks (2003), consumers who have lodged a
with the results of Study 1. complaint typically want to see some form of correction. Previous
research in social psychology and consumer behavior (Elliot, 2006;
3.1.1. Service recovery actions Hui and Bateson, 1991; Roth and Cohen, 1986) offers two possible
3.1.1.1. Compensation. Consumers evaluate the fairness of a corrective actions: approach or avoidance. Service providers may
service recovery along three dimensions: distributive justice, pro- put an end to a service problem (approach strategy). For instance,
cedural justice, and interactional justice (Smith et al., 1999). service employees or managers could stop smokers from lighting
Compensation is often associated with restoring customers’ per- up cigarettes in a restaurant. If so, the cause of the problem is
ceptions of distributive justice. Specifically, compensation in the entirely eliminated.
form of a discount or a refund is viewed as tangible redress for a However, when direct confrontation is not feasible, the service
customer’s economic loss (e.g., time and money). Compensation provider might opt for an indirect action. For example, the service
has been demonstrated as an effective recovery effort to increase provider might remove the complaining customer from the source
satisfaction and repurchase intentions (e.g., Boshoff, 1997; Grewal of problem (avoidance strategy). This approach-avoidance action
et al., 2008; Mount and Mattila, 2000) or to reduce negative word is similar to consumer responses to service failures as consumers
of mouth (Maxham, 2001; Wirtz and Chew, 2002). can complain directly (approach) to the service provider about
their dissatisfaction or remain silent (avoidance). Compatibility
management also suggests that service providers should separate
3.1.1.2. Apology. An apology is viewed as a valuable reward that
customers from any disturbances such as noisy children (Huang,
reinstates esteem in an exchange relationship (Walster et al., 1973).
2008; Martin and Pranter, 1989).
Offering an apology indicates that an organization accepts the
Based on the above discussions, it is logical to predict that:
responsibility for the negative event (Coombs, 2007). Previous
research has shown that an apology not only enhances customers’
perceived distributive justice (Tax et al., 1998) but also interac-
tional justice (Blodgett et al., 1997; Goodwin and Ross, 1992; Smith H1. Compared to other recovery methods, confronting the deviant
et al., 1999). Empirical evidence also shows a significant impact of customer will receive the highest ratings in terms of (a) perceived
apology on satisfaction, word-of-mouth, and repurchase intentions employee effort, (b) complaint handling satisfaction, and (c) behav-
(e.g., Davidow, 2000; Wirtz and Mattila, 2004; Wong, 2004). ioral intention.
184 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192

The effectiveness of an apology and compensation is undis- H4. Gender moderates the effectiveness of various recovery
putable in the East Asian culture. For instance, Wong (2004) actions on (a) perceived employee effort, (b) complaint handling
demonstrated that compensation improved customers’ assess- satisfaction, and (c) behavioral intention.
ments of the service encounter in the US, Australia, and Singapore,
while apology improved satisfaction for the Singaporean and Aus- 3.2. Materials and methods
tralian samples only. The relative effectiveness of an apology
and compensation, however, remains unknown. Furthermore, a A 6 (recovery actions: no action, apology, compensation, reloca-
removal of the focal customer as a recovery strategy has not tion, confrontation, control) × 2(gender: male vs. female) between
been empirically examined. Hence, we propose the following non- subjects design was employed. The use of a scenario method is
directional hypothesis: supported by Weiner (2000) and Goodwin and Ross (1992) as
this method avoids the problems of intentionally imposing service
H2. There are significant differences among apology, compen- failures on customers and it minimizes memory-bias, which is com-
sation, and relocation of the focal customer with regards to (a) mon in self-reports of service failures (Smith et al., 1999).
perceived employee effort, (c) complaint handling satisfaction, and
(c) behavioral intention. 3.2.1. Scenario development
Based on the results of Study 1, a restaurant context was chosen.
The failure scenario described an incident where a fellow customer
3.1.2. Baseline behavior intention smoked in a non-smoking area of a casual dining restaurant. The
In the context of deviant customer behaviors, upon experi- recovery actions taken by the service staff were (a) no action, (b)
encing the unpleasant consumption situation, customers tend to a sincere apology, (c) compensation in terms of 15% discount, (d)
exhibit certain behavioral intentions. This baseline intention might relocation to another table further away from the smokers, and (e)
be attenuated or ameliorated depending on subsequent recovery confronting the deviant customer to stop the deviant behavior. A
actions. In this study, we used a control group as a baseline mea- control scenario was also introduced as a baseline for the behavioral
sure for behavioral intention (i.e., behavioral intent without any intention measure. A sample of scenario is shown in Appendix A.
complaints). Service recovery paradox (Smith and Bolton, 1998)
suggests that customer satisfaction and repurchase intention are 3.2.2. Research procedure and measurements
higher with successful recovery actions than if no failure occurred Subjects were adult consumers recruited through a mall inter-
in the first place. Voorhees et al. (2006) also demonstrated that cept study from a large shopping center in the Klang Valley area
repurchase intention among non-complainants is higher than cus- in Malaysia. Systematic random sampling with an interval of 25
tomers who complain and receive no recovery or dissatisfactory was employed. The recruitment process covered weekdays and
recovery effort. On the other hand, non-complainants exhibit lower weekends, as well as different times of a day to minimize bias. Sub-
repurchase intention than customers who experience satisfactory jects were asked to read one of the six scenarios and to indicate (i)
recoveries. In a similar vein, we put forth the following hypothesis: the realism of the scenario, (ii) the severity of the service failure,
(iii) their negative emotions toward the situation—other customer
H3. Customers’ behavioral intentions will be (a) higher for recov- smoking in a non-smoking area, (iv) the perceived controllability
ery actions but (b) lower for no action compared to control. of the deviant customer behavior, and (v) their thoughts pertaining
to the scenario described. Scenario realism and severity of service
failure were adapted from and Goodwin and Ross (1992), the nega-
3.1.3. Gender differences tive emotions was captured by scales developed by Patterson et al.
Past research has documented that males and females exhibit (2006) while the controllability item was modified from Wirtz and
different traits (e.g., Allinson and Hayes, 1996; Iacobucci and Mattila (2004) for the context of deviant customer behavior.
Ostrom, 1993). Overall, females are perceived as more nurtur- Except for the control scenario, perceived employee effort was
ing and yielding while males are more assertive and instrumental measured using items from Mohr and Bitner (1995), while com-
(Iacobucci and Ostrom, 1993). In addition, males tend to engage in plaint handling satisfaction was assessed using scales from Mattila
more analytic and selective information processing while females and Patterson (2004). Behavioral intention measures were adapted
are generally more subjective and comprehensive in processing from Bougie et al. (2003) and Wirtz and Mattila (2004) and were
information (Allinson and Hayes, 1996; Darley and Smith, 1995). employed in all six scenarios. Appendix B presents these measure-
In the context of complaint handling, Hess et al. (2003) pointed ment scales and scale reliabilities.
out that female customers hold higher expectations toward service
recovery than their male counterparts. Moreover, women appear 3.3. Results
to put more focus on the process than men, whereas men tend to be
more task focused and thus emphasize the outcome (e.g., Iacobucci A total of 188 females and 189 males spanning the three major
and Ostrom, 1993; Mattila et al., 2003). As a result, a service ethnic groups in Malaysia participated in this study. The chi-square
employee needs to be an active listener and invite customer input analyses indicated no significant differences in the association
to effective recovery when handling female complainants (Gruber between the 12 treatment groups and ethnicity (2 = 12.16, p = .95),
et al., 2009; McColl-Kennedy et al., 2003). On the other hand, men education level (2 = 21.96, p = .93), and occupation (2 = 55.43,
expect speedy recovery which helps them to save time (Gruber p = .12). However, the association between the treatments and prior
et al., 2009). In other words, female customers emphasize the pro- work experience in a restaurant (2 = 21.58, p = .03) and marital
cess while males concentrate on the outcome. Mattila et al. (2009) status (2 = 24.67, p = .01) were significant.
also demonstrated that while both apology and compensation Eighty-eight percent of the respondents had at least a diploma
recovery actions improved satisfaction among male respondents, (similar to some college degree in the US) while 28.4% held a
a combination of apology and compensation is needed to induce managerial or administrative position and 17.8% worked in a
above neutral post recovery satisfaction among female respon- professional or technical field. Almost half (41.8%) of the male
dents. Therefore, we propose that gender moderates the impact of respondents had prior work experience in the restaurant indus-
recovery efforts on customer outcomes. More specifically, we put try but only 23.9% among the females. Also, the majority of the
forth the following prediction: male respondents were married (52.9%) while females were largely
 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192 185

Table 4
Scenario realism, severity of customer misbehavior, negative emotion, and perceived controllability.

Treatment Realism Severity Negative emotion Controllability

Female
Control (n = 35) 5.80 ± 0.15 6.06 ± 0.15 6.17 ± 0.14 5.86 ± 0.21
No action (n = 31) 5.81 ± 0.18 6.23 ± 0.16 6.23 ± 0.14 6.06 ± 0.15
Apology (n = 30) 5.87 ± 0.19 6.40 ± 0.13 6.08 ± 0.16 6.10 ± 0.23
Compensation (n = 32) 5.75 ± 0.17 6.41 ± 0.16 5.86 ± 0.15 5.88 ± 0.25
Relocation (n = 30) 5.70 ± 0.18 6.43 ± 0.20 6.05 ± 0.18 5.87 ± 0.29
Problem resolution (n = 30) 5.73 ± 0.19 5.97 ± 0.21 5.82 ± 0.15 5.43 ± 0.28
Male
Control (n = 34) 5.59 ± 0.19 6.24 ± 0.15 5.85 ± 0.16 6.06 ± 0.18
No action (n = 30) 5.70 ± 0.18 6.50 ± 0.11 6.08 ± 0.12 6.07 ± 0.22
Apology (n = 30) 5.60 ± 0.16 6.10 ± 0.16 6.08 ± 0.15 5.23 ± 0.34
Compensation (n = 32) 5.75 ± 0.16 6.41 ± 0.14 6.09 ± 0.15 5.56 ± 0.24
Relocation (n = 32) 6.00 ± 0.14 6.34 ± 0.13 6.25 ± 0.14 5.69 ± 0.31
Problem resolution (n = 31) 5.87 ± 0.17 6.00 ± 0.17 5.89 ± 0.15 6.03 ± 0.21

Note. Values presented are mean ± SE. Realism was measured with 7-point scale (1 = highly unrealistic, 7 = highly realistic); severity was measured with 7-point scale (1 = not
at all, 7 = very much so); negative emotion was measured via two items—angry and annoyed/irritated (1 = not at all, 7 = very much so); controllability was measured with
7-point Likert scale (1 = strongly disagree, 7 = strongly agree).

single (67%). However, there was no significant difference on the and n = 3 female) in comparison to problem being resolved in the
average age (ranging from 28.3 to 33.2 with mean 30.6 years old) confrontation scenario (35.5% males and 43.3% females).
of the respondents among the treatment groups (F = 0.54, p = .88). To further substantiate the insignificant effect of apology in the
compensation, relocation, and confrontation scenarios, a thought
protocol analysis on apology and no action scenarios was also con-
ducted. Results for the context of no action showed that only one
3.3.1. Manipulation checks female indicated the neglect of apology while the majority of the
The manipulation checks indicated that the stimuli were effec- subjects expected service staff be more responsible to uphold the
tively executed (Table 4). The scenarios were rated as highly restaurant policy and to resolve the problem (36.7% male and 63.3%
realistic (mean ratings varied from 5.59 to 6.00, F = .46, p = .93) females). More importantly, both genders indicated their expecta-
and respondents believed that the restaurant had control over tions of further action be taken in the mere apology scenario (56.7%
the deviant customer behavior (mean scores ranged from 5.23 to males and 86.7% females). Taken together, it is reasonable to con-
6.10, F = 1.27, p = .24). The severity of the deviant behavior varied clude that the confounding effect of apology is not likely to be a
from 5.97 to 6.50 (F = 1.35, p = .19). Negative emotions (Cronbach’s threat in the present study and the operationalization of compen-
˛ = .79) toward the deviant behavior were consistently high across sation and relocation was acceptable. The confrontation scenario is
the treatment groups (mean scores ranged from 5.82 to 6.23, relabeled as problem resolution subsequently.
F = 1.02, p = .43).
Due to the significant associations between the treatment
3.3.2. Hypotheses testing
groups and both marital status and prior restaurant work expe-
The scale reliability coefficients of perceived employee effort,
rience, further analyses were conducted (Perdue and Summers,
complaint handling satisfaction, and behavioral intention were
1986; Voorhees et al., 2006). Multivariate analysis of variance
0.86, 0.97, and 0.81, respectively. A MANOVA approach was
(MANOVA) was applied to control for the correlation between
employed to control for the experimental-wise error rate
perceived employee effort and complaint handling satisfaction
(Tabachnik and Fidell, 1996). Behavioral intention was subject to
(r = .84). Results show that both marital status (Wilks’s  = .98,
univariate analysis due to the presence of a control group.
F = 2.71, partial 2 = .02, p > .05) and prior restaurant work expe-
The results of the MANOVA table indicate a significant inter-
rience (Wilks’s  = 1.00, F = 0.22, partial 2 = .00, p > .05) did not
action effect (Wilks’s  = .91, F = 3.45, p < .01) and a main effect for
significantly affect the two dependent variables. A univarite anal-
recovery action type (Wilks’s  = 3.70, F = 47.89, p < .01). The main
ysis of variance (ANOVA) was conducted on behavioral intention.
effect of gender was insignificant (Wilks’s  = .98, F = 2.60, p = .08).
Likewise, marital status (F = 1.50, partial 2 = .00, p > .05) and prior
To determine which variables cause differences among the sce-
restaurant work experience (F = 0.16, partial 2 = .00, p > .05) did not
narios, independent univariate tests with post hoc analysis were
exert significant effect, thus they were dropped from the subse-
performed. The results of analysis for the three dependent variables
quent analyses.
are provided in Table 5.
The current study also analyzed the operationalization of the
recovery actions via customers’ thought protocol. The potential
confounding effect of apology, especially in the compensation, relo- 3.3.2.1. Perceived employee effort. Results of Tukey’s multiple com-
cation and confrontation scenarios needs to be addressed. The parison tests provide support for H1a. Resolving deviant customer
analysis of the thought protocols was executed in a similar manner behavior (M = 5.66, SE = 0.12) resulted in higher effort ratings than
to that in Study 1. The inter-judge agreement was 0.79, exceeding any of the other recovery actions (p < .05): apology (M = 3.00,
the 0.59 required proportion (Rust and Cooil, 1994). Disagreements SE = 0.14), relocation (M = 3.78, SE = .19), compensation (M = 4.30,
were resolved through discussion. SE = 0.15), and no action (M = 2.06, SE = 0.13). However, H2a
The results indicate that apology was only occasionally men- received mixed results. Compensation was not significantly differ-
tioned by both male and female customers in the compensation ent from relocation (p = .19) but both were rated higher in perceived
(n = 1 for male and n = 4 for female) while compensation was noted employee effort than apology (p < .05).
by 43.8% males and 46.9% females. Similarly, apology was only cited To examine H4a, a comparison of the recovery actions was
by a few respondents (n = 2 for male and n = 1 for female) compared performed separately for male and female consumers. Perceived
with the intended relocation suggestion (53.1% male and 53.3% employee effort ratings for women were as followed: resolving
female). Finally, apology was also rarely commented (n = 2 for male deviant customer behavior (M = 5.27, SE = 0.16), relocating the focal
186 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192

Table 5
Consumer thoughts analysis for male and female participants in five scenarios.

Female N Scenario N Male

31* No actiona 30*

Very angry and dissatisfied with the response from the server 10 General negative 15 I will be angry with the attitude of the server and will scold
response him/her
The server doesn’t care about the problem and is irresponsible. I will leave the restaurant
Perhaps he doesn’t want to get into trouble but to let the
customers continue smoking in the non-smoking area
Angry, the server should do something to solve the problem 20 Problem solving 11 I will try to complain to the manager and get the solution
expectation immediately
I think the server should ask the customers to stop smoking or I will obviously feel angry because the server should at least
request them to change to the smoking area approach the group of smokers and stop them
They should stick to their “no smoking” policy by asking the Sincere it is non-smoking area, it is better to stop the smokers
customers to move to the smoking area than do nothing. If the restaurant policy is useless, customers
will not visit again
Charge the smoker
I am very unhappy with the response of the server. The server 1 Apology expectation NA
didn’t try to solve my problem. He didn’t even apologize
I will be very angry and immediately change my seat 1 Self action 4 Angry. I may stop the smokers myself. I will scold the smokers.
I will relocate the table
30* Mere Apologyb 30*
Angry and disappointed 3 General negative 7 Totally disappointed with their poor customer service
response
Perhaps the server is afraid of the four tough looking men. She 1 Understanding/satisfied 2 I understand that the server could only apologize but do
could only apologize to me but not stop the smokers with apology nothing as “customer is always right”. They cannot offend their
customers. Anyway, maybe the server can relocate me to a
better place, away from the smokers
I am satisfied because the server apologized to me
Apology is a must but I am still irritated because the smokers 26 Inadequacy of mere 17 The server was polite but my problem remained unresolved
are still smoking apology
After apologized, the server should try to approach the The server should ask the smokers to stop smoking, not just
smokers to remind them that they are in non-smoking area apology. I can’t accept it
Although the server apologizes to me, I think he/she should do Must obey the policy to stop smoking. Apology is not enough
something to avoid this situation from happening
The servers should at least offer a better service such as give
discount or coupon in return
I will feel angry and request to switch place
NA Self action 1 I will ask the customer to stop smoking and will not visit again
32* Compensationb 32*
Satisfied with the way the server handled the complaint 5 General negative 2 Not bad
response
I am satisfied with the apology that was given by the service 4 Satisfied with apology 1 Apologize is acceptable but instead of giving discount, the
crew yet I am not comfortable to dine in this restaurant server should tell the smoker about the non-smoking area
because the service crew didn’t stop/warn the customers
who smoke in a non-smoking area
Accept the discount given 3 Positive attitude NA
toward compensation
I think the treatment (15% discount) is fair. I am quite happy
with it
The server should ask the customers who smoke to switch to 2 Negative attitude 4 I will be very angry. The server didn’t enforce the policy. I will
the smoking area but there is no need to give me toward compensation not take the discount because it spoils my day
compensation. Compensation is not important for me
I don’t think discount can compensate my loss
Should stop smoking. Discount is not enough
The server should politely ask the smoking customer to move 10 Compensation is fine 10 I am pleased to be offered a discount but I would be more
to the smoking section. Compensating other customer is but not inadequate pleased if the server could ask the four men to stop smoking or
very welcomed due to the failure of the server move to smoking area
Dissatisfied and unhappy. I expect the server approach the 8 Other action 9 Annoyed. The restaurant should not allow customer smoke in
smokers and ask them to move to the smoking area politely non-smoking area
and firmly
The action was right but not good enough because they did not
bring the four tough-looking smokers to smoking area.
Discount is fine but the problem still needs to be solved
30* Relocationa 32*
Annoyed, angry, unfair to other customers 3 General negative 7 Disappointed. Without hesitation, I will visit again
response
Fail to serve customer with respect and can’t solve very simple
situation
It is a good response from the server. I would be pleased with 5 General positive NA
the service response
I feel that this is the only solution the server could offer to me
since I think that the major problem is the attitude of the
smokers
Although it is not the fault of the restaurant, the server 1 Satisfied with apology 2 Although the apology from the server is acceptable, the server
apologized to us. I am satisfied. The smokers should not sit in should stop the smoking instead of switch my seat
non-smoking area
 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192 187

Table 5 (Continued)

Female N Scenario N Male

Not satisfied. The customers who smoked should be relocated 9 Challenged/negative 17 The action was unfair. Why should I move to another table?
toward relocation
suggestions
I will be angry because the server should relocate the smoker
not me
Can’t accept. It is the smokers’ mistake. So it should be them
who need to be relocated. I am eating now, so it is impossible
for me to move to another area. The server should ask them to
go outside to smoke and tell them to respect other customers
It is good that the management seated me in another area 7 Positive attitude NA
because for sure I, as a woman, don’t like smokers toward relocation
Quite satisfied! Since the server try to solve my problem, I am
willing to shift to a “smoke free” zone
Angry because the server did not ask the smokers to leave 8 Other action 13 The server should ask the smokers to stop smoking instead of
relocation
I will be annoyed by the customers. But the more frustrating is Disagree with the relocation. They are supposed to ask the
being told by the server to change my seat to another table smokers to stop smoking
away from the smokers
30* Problem resolutionb 31*
I am satisfied. The best action to handle the situation 5 General positive 11 Their action makes me feel good
response
The server did her best Satisfied
The smokers are annoying. After server apologizes, I am happy 3 Satisfied with apology 2 Of course the server has to apologize and talk to the smokers
politely
This restaurant gives me good service when they solve the 13 Problem resolved 11 The server tries to reflect my problem to others and so I am
problem quickly delight on the action of the server
I am happy with the way the server solve my problem Feel appreciated and the problem addressed.
Even though it was rectified, it should not have happened 8 Other action 7 Acceptable but the restaurant should enforce the policy
The server should notice and act immediately. He/She should Satisfied, but I think the server should stop the smokers before
not wait until I complaint customer complain
I think the server should stop the smokers before they even
begin and charge those who have violated the policy
*
Represents the total number of respondents.
a
The sum is more than the total because of multiple comments.
b
The sum is less than the total because some respondents did not comment.

customer (M = 4.31, SE = 0.28), offering compensation (M = 4.21, rating among female respondents (p < 05). This was followed by
SE = 0.23), apologizing (M = 2.99, SE = 0.17), and no action (M = 2.41, relocation of the focal customer (M = 3.82, SE = 0.31) and compen-
SE = 0.19). Resolving deviant customer behavior led to the high- sation (M = 3.66, SE = 0.27, p > .05). Again, an apology (M = 2.07,
est ratings of effort (p < .05). Offering compensation was viewed as SE = 0.16) was just as good as no action (M = 1.59, SE = 0.14, p > .05).
equally effortful as relocating the focal customer (p > .05). Mean- Both of these recovery actions were the lowest in complaint
while, a mere apology was indifferent from no action (p > .05) and handling satisfaction (p < .05).
both received the lowest perceived effort ratings (p < .05). For male respondents, the pattern was slightly different. Con-
The results for male respondents showed a somewhat different fronting the deviant customers to resolve the deviant behavior
pattern: deviant customer behavior resolution (M = 6.03, SE = 0.16), (M = 6.17, SE = 0.13) resulted in highest satisfaction with complaint
compensation (M = 4.39, SE = 0.18), relocation (M = 3.28, SE = .24), handling (p < .05). Unlike female respondents, offering compensa-
apology (M = 3.01, SE = 0.23), and no action (M = 1.69, SE = 0.15). tion (M = 4.01, SE = 0.23) to male customers received significantly
Resolving deviant customer behavior was viewed as exhibiting the higher satisfaction rating than relocating them (M = 2.95, SE = 0.24,
greatest amount of employee effort (p < .05). Unlike female respon- p < .05). Relocating the focal male customers was not different from
dents, relocating the focal customer was perceived as less effortful
than offering compensation (p < .05), but it was not different from
an apology (p > .05). No action received the lowest perceived effort 7

rating (p < .05). Hence, H4a was supported. Comparison between

6
Perceived Employee Effort

male and female respondents across the different recovery action

also revealed that women viewed relocation as significantly higher
5
in effort than their male counterparts (p < .05) (Fig. 1).
4
3.3.2.2. Complaint handling satisfaction. The results of a planned
comparison analysis for complaint handling satisfaction are visu- 3
alized in Fig. 2. As with perceived employee effort, apology
(M = 2.27, SE = .14), relocation (M = 3.37, SE = .20), compensation 2
(M = 3.83, SE = .18), and no action (M = 1.49, SE = .09) were all rated
lower (p < .05) than resolving deviant customer behavior (M = 5.78, 1
SE = .14). Again, the results for H2b were mixed. Compensation was No Acon Apology Relocaon Compensaon Problem
Resoluon
not different from relocation (p > .05) but both were rated higher in
complaint handling satisfaction than apology (p < .05). Female Male
We then examined the effectiveness of the recovery actions
Fig. 1. Perceived employee effort of five recovery actions for male and female note.
(H4b) by gender. The results show that resolving deviant customer Perceived employee effort was measured with 7-point Likert scale (1 = strongly dis-
behavior (M = 5.37, SE = 0.23) received the highest satisfaction agree, 7 = strongly agree).
188 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192

7 (p > .05). Thus, H3a was only partially supported. In addition, no

Complaint Handling Satisfaction

action resulted in lower behavioral intention than the control

6 condition (p < .05) substantiating H3b.
In terms of gender, behavioral intention ratings for women
5 were as followed: resolving deviant customer behavior (M = 4.98,
SE = 0.22), compensation (M = 4.22, SE = 0.16), control (M = 4.13,
4
SE = 0.22), relocating the focal customer (M = 4.07, SE = 0.25), apol-
ogy (M = 3.07, SE = 0.15), and no action (M = 2.98, SE = 0.19). Among
3
female respondents, resolving deviant customer behavior, offering
compensation, and relocating the focal customer resulted in simi-
2
lar levels of behavioral intention as the control condition (p > .05).
Both no action and a mere apology received the lowest behavioral
1
No Acon Apology Relocaon Compensaon Problem intention ratings (p < .05).
Resoluon A somewhat distinct pattern of results was found for male
Female Male respondents. Resolving deviant customer behavior (M = 5.59,
SE = 0.23) led to the highest level of behavioral intention (p < .05).
Fig. 2. Complaint handling satisfaction of five recovery actions for male and female
Compensation (M = 4.02, SE = 0.16), relocation (M = 3.67, SE = 0.18),
note. Complaint handling satisfaction was measured with 7-point Likert scale
(1 = strongly disagree, 7 = strongly agree). and apology (M = 3.74, SE = 0.15) were not significantly different
(p > .05) from the control condition (M = 3.92, SE = 0.17). No action
(M = 2.48, SE = 0.22) received the lowest behavioral intention rating
a mere apology (M = 2.47, SE = 0.24, p > .05). No action (M = 1.38,
(all p < .05). Hence, H4c was confirmed.
SE = 0.10) was rated lowest in complaint handling satisfaction
(p < .05). Overall, H4b was substantiated. A comparison between
4. General discussion
male and female respondents across the different recovery actions
also revealed that women were significantly more satisfied with
Prior research suggests that deviant customer behaviors are
relocation than their male counterparts (p < .05).
ubiquitous and a common cause of dissatisfaction. The present
study demonstrates the impact of deviant customer behaviors in
3.3.2.3. Behavioral intention. A univariate ANOVA was performed an East-Asian context—Malaysia, the effectiveness of various recov-
on behavioral intention due to the presence of a control group ery actions, and gender differences in consumers’ reactions to such
(Fig. 3). The results yielded a significant interaction effect (F = 3.37, actions. We introduce approach and avoidance strategies to the
p < .01) and main effect for recovery action type (F = 37.24, p < .01) battery of potential remedies when dealing with deviant customer
while the main effect for gender was insignificant (F = .00, p = .90). behaviors (Boshoff, 1997; Levesque and McDougall, 2000; Wirtz
Solving deviant customers behavior (M = 5.29, SE = 0.16) was and Mattila, 2004).
rated higher than apology (M = 3.41, SE = 0.12, p < .05), reloca- The findings from Study 1 indicate that Malaysian consumers
tion (M = 3.86, SE = 0.15, p < .05), compensation (M = 4.12, SE = 0.11, are not likely to voice their dissatisfaction about deviant customer
p < .05), and no action (M = 2.73, SE = 0.15, p < .05), hence supporting behaviors. It is also intriguing to see that they expect the service
H1c. The results for H2c were mixed. Specifically, compensation provider to take action to rectify the unpleasant situation. One plau-
was not statistically different from relocation (p > .05) but induced sible explanation is that many deviant customer behaviors may be
greater levels of behavioral intention compared to apology (p < .05). deemed controllable by the service provider (Huang, 2008). This
On the other hand, relocation was not significantly different from finding is also in line with the notion of collectivism and a face con-
that of an apology (p > .05). cern associated with East-Asian cultures. There are two types of
The purpose of H3a and H3b was to compare the recovery face concern, namely self-face and other face (Ting-Toomey et al.,
actions with the baseline (control situation). The results indicate 1991). The salience of self-face concern refrains consumers from
that resolving deviant customer behavior led to the highest complaining when the complaint situation involves a direct con-
behavioral intention (p < .05) while apology received a lower rating frontation with the responsible party (Chiu et al., 1988). On the
(p < .05). On the other hand, relocation and compensation did not other hand, collectivists are also more likely to preserve other-face
result in higher behavioral intention than the control condition (i.e., not embarrassing others publicly), hence leading to avoidance,
integration, and compromise strategies (Brew and Cairns, 2004;
6 Ting-Toomey et al., 1991).
5.5 Study 2 revealed that an employee confronting the deviant cus-
tomer to stop the deviant behavior (i.e., smoking in a non-smoking
5
Behavioural Intention

area of a restaurant) generally led to the highest levels of perceived

4.5
effort and satisfaction with complaint handling. This is not surpris-
4
ing since customers typically expect the problem to be resolved
3.5 when they voice their concerns. The significance of resolving the
3 problem on the spot is in line with prior studies in relationship
2.5 marketing (e.g., Andreassen, 2000; Heung and Lam, 2003). In a sim-
2 ilar vein, not taking action upon customer complaints leads to low
1.5
ratings of perceived effort and complaint handling satisfaction.
However, these findings for resolving deviant customer behav-
1
Control No Action Apology Relocation Compensation Problem ior or not taking action did not carry over to behavioral intention.
Resolution It is interesting to note that offering compensation and relocating
Female Male the focal female customer resulted in similar levels of behavioral
intention as resolving the deviant customer behavior. There was
Fig. 3. Behavioral intention of five recovery actions and a control for male and female
note. Behavioral intention was measured with 7-point Likert scale (1 = strongly dis- no difference between an apology and no action among the female
agree, 7 = strongly agree). participants. Conversely, compensation, relocation, and apology led
 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192 189

Table 6
Multivariate and univariate analysis of variance results.

Multivariate results Univariate results

PEE CHS BI

Wilk’s  F Partial 2
F Partial 
2
F Partial 
2
F Partial 2

Main effect
Gender (G) 0.98 2.60 0.02 1.50 0.01 0.47 0.01 0.01 0.01
Recovery action (RA) 3.70 47.89** 0.39 87.42** 0.54 112.80** 0.60 37.24** 0.34
Interaction
G × RA 0.91 3.45** 0.04 6.19** 0.08 4.51** 0.06 3.37** 0.04

Note. PEE = perceived employee effort; CHS = complaint handling satisfaction; BI = behavioral intention.
**
p < .01.

to similar levels of behavioral intent among males. There are two Finally, the current study included a control group (i.e., no
possible explanations to these findings: satisfaction-behavioral complaint voiced) as the baseline for behavioral intention. It is
intention link and globality attribution. important to note that no action was rated significantly lower than
Past research examining customer switching behaviors points the baseline across the two genders. This finding parallels the work
to the critical role of switching costs. In particular, switching costs by Voorhees et al. (2006) suggesting that non-complainers are more
were found to interact with customers’ satisfaction on their repur- likely to repurchase than customers who complained but did not
chase and word-of-mouth intention (e.g., Han et al., 2009; Lam receive any recovery action. In sum, the results of this study high-
et al., 2004). The positive relationship between satisfaction and light the importance of taking action when the failure is caused by
behavioral intention was inhibited when switching costs were high. deviant customer behaviors.
Besides switching barriers, attribution of globality (Abramson
et al., 1978), which refers to the customers’ perceptions whether 5. Conclusions
the cause of an event occurs in multiple settings, as opposed to
just the focal one, provides another viewpoint to the satisfaction- This research explores common deviant customer behaviors in
behavioral intention relationship. Huang et al. (2010) show that a an East-Asian service context—Malaysia. To the best of our knowl-
globality attribution buffers the negative impact of controllability edge, this study is the first to include the approach-avoidance
and stability attributions on customer satisfaction and behavioral strategy to the battery of service recovery actions. The effectiveness
intentions. Translating into our context, customers may perceive of different recovery actions on perceived effort, complaint hand-
that the deviant customer behavior is so pervasive and will occur ling satisfaction, and behavioral intention in a restaurant setting
in other service establishments that it is pointless to switch. This was investigated in light of the role of gender. Our study findings
explanation, nevertheless, warrants future empirical investigation. have several important managerial implications.
Although we found some gender differences in consumers’ reac-
tions to various recovery actions, a word of caution is needed here. 5.1. Managerial implications
The gender difference might be limited to the relocation scenario.
Overall, the effectiveness of each recovery action was generally First, confronting the deviant customer is the best course of
uniform across the two genders except for relocation. Specifically, action. It is therefore imperative that service firms design train-
men were not as appreciative of relocation as a recovery effort. ing programs to help frontline employees in dealing with deviant
The thought protocol analysis (Table 6) confirmed that male cus- customers. In addition, employees should be trained to be sensi-
tomers were more likely to oppose the relocation strategy than tive to customer dissatisfaction caused by other customers’ deviant
their female counterparts (53.1% males vs. 30% females). More- behaviors.
over, none of the male respondents accepted the relocation solution Second, employees should be aware of the different expecta-
while 23.3% females did. In fact, male customers equated such an tions between male and female customers. Specifically, compen-
action with a mere apology. None of the male subjects in the relo- sation seems to be the winning strategy among male customers.
cation scenario indicated their satisfaction toward the suggestion Conversely, relocating female customers away from the deviant
and only two were happy in the mere apology scenario. customer might be a cost-effective recovery strategy. Service
The current study also revealed that apology is a necessary yet employees should be reminded that a simple apology is insufficient.
an insufficient condition to restore negative experiences, in par- However, this does not, in any way, suggest that service employees
ticular among women. An apology needs to be followed by other should not apologize. In fact, apologizing is particularly important
actions. On the whole, these differences between male and female for Chinese consumers in Malaysia (Boo et al., 2008). An apology
customers are consistent with past research (McColl-Kennedy should be used together with other types of recovery actions.
et al., 2003; Gruber et al., 2009) and underscore the importance of Finally, the results of this study highlight the importance of
offering various recovery options based on the customer’s gender. compatibility management (Martin and Pranter, 1989). Service
In particular, prior research shows that men tend to be more providers have several means to curb deviant customer behaviors.
aggressive, more competitive, more willing to take risk, and more For instance, restaurant firms might want pay attention to their ser-
outcome-focused (Leventhal and Garcia, 1991; McColl-Kennedy vicescape design to prevent potential customer deviant behaviors
et al., 2003). Thus, they might aggressively challenge the reloca- (Miao et al., 2011). Moreover, educating the customer about codes
tion strategy. On the contrary, the nurturing, accommodating, and of conduct might reduce deviant customer behaviors.
empathetic characteristics of women (Becker et al., 2002; Iacobucci
and Ostrom, 1993; Meyers-Levy, 1988) may lead them to value 5.2. Limitations and future research
such a recovery effort more. Furthermore, men expect a speedy
resolution in order to save time and focus on other activities There remain numerous opportunities for further research. The
(Gruber et al., 2009). As a result, being asked to move to another present research focused on a single type of deviant customer
section of the restaurant might be perceived as cumbersome while behavior, namely smoking in non-designated areas, in an East-
compensation is perceived as a valuable recovery strategy. Asian culture (i.e., Malaysia). Consumers view this type of deviant
190 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192

customer behavior as controllable by the service provider, and A.4. Compensation scenario
hence some form of recovery action is expected. Future studies
might want to examine misbehaviors that are beyond the control (Continued from control scenario) You call for the server who
of the service provider, such as a baby crying or customers making approaches you immediately. You complain about the smoking cus-
excessive noise while celebrating. A cross cultural study would be tomers. The server sincerely apologizes to you and then offers a
a fruitful avenue in understanding the effectiveness of relocation, compensation of 15% discount to your bill.
a recovery action which is associated with the face concern in a
collectivist eastern culture. A.5. Relocation scenario
The conceptualization of problem resolution warrants further
research. Given that less than half of the respondents regarded the (Continued from control scenario) You call for the server who
scenario as problem resolution, this recovery effort might induce approaches you immediately. You complain about the smoking cus-
multiple interpretations. For instance, it may be viewed as a tem- tomers. The server sincerely apologizes and asks if you would like
poral solution. A genuine solution would be upholding the policy at to change to another table further away from the smokers.
all times and proactively preventing deviant customer behaviors.
It is also important to note that the manipulation checks on the A.6. Problem resolution scenario
recovery actions were not measured but based upon consumers’
thought listings. The no action scenario might be judged as rude. (Continued from control scenario) You call for the server who
The apology scenario might also be perceived as mere apology approaches you immediately. You complain about the smoking cus-
without action since the thought protocol revealed that consumers tomers. The server sincerely apologizes to you and then approaches
expected further actions. Hence, future study may explicitly mea- the customers. Politely the server advises the customers to stop
sure the manipulation of the actions. smoking and the customers discontinue smoking.
In addition, other customers’ deviant behaviors may be asso-
ciated with different types of losses. For instance, breaking in line Appendix B. Measurement items and reliability (Study 2)
may lead to a service delay (time loss) while dining with friends in a
noisy environment may cause emotional stress (emotional loss) or Measurement items
embarrassment (social loss). Future research might want to exam-
Realism of scenario
ine the impact of such losses on consumers’ fairness perceptions. How realistic do you think is the Highly unrealistic (1)/highly
Finally, the lack of transferability of the results from perceived problem (other customers smoking) realistic (7)
effort and complaint handling satisfaction to behavioral intention given in the scenario?
was suggested to be caused by switching barriers and attribution Severity of service failure
of globality. Further empirical research is needed to examine the How important is the service failure Not at all (1)/very much so (7)
effect of these two factors. (other customers smoking)?
Negative emotion (Cronbach’s ˛ = 0.79)

How angry are you with the service Not at all (1)/very much so (7)
Appendix A. Scenario descriptions failure? (other customers smoking)
How annoyed/irritated are you with
the service failure? (other customers
A.1. Control scenario smoking)
Controllability
On a Saturday evening, you and your friends have decided to go The cause of customers smoking in Strongly disagree (1)/strongly
the non-smoking area is controllable agree (7)
out for a relaxing dinner in the casual dining restaurant in town.
by this restaurant
This is the first time you visit the restaurant. You enter the restau- Perceived employee effort (Cronbach’s ˛ = 0.86)
rant and request for a table at the non-smoking area. The hostess
The server seems to care about my Strongly disagree (1)/strongly
promptly seats your party at a table in the non-smoking section of
problem agree (7)
the dining room. After placing your orders, you look around, enjoy The server does not spend much time
the music and the view of the city. You find the atmosphere in the in taking care of the problem (R)
restaurant to be a pleasant blend of comfort and tranquillity. I perceive the server has put in a lot
Within 20 min, your meal is served. While you are enjoying the of effort into serving me
The server really tries to solve my
delicious food, four tough-looking men walk in and sit at the table problem
next to yours. After placing their orders, the men light up their Complaint handling satisfaction (Cronbach’s ˛ = 0.97)
cigarettes. The fumes float over to your table.
I am satisfied with the restaurant on Strongly disagree (1)/strongly
the resolution to my complaint agree (7)
I am satisfied with the way my
A.2. No action scenario complaint was handled
I am satisfied with the restaurant in
this particular dining experience
(Continued from control scenario) You call for the server who
Behavioral intention (Cronbach’s ˛ = 0.81)
approaches you immediately. You complain about the smoking
customers. The server just smiles, shrugs his shoulders (a sign of How likely are you to visit the Very unlikely (1)/very likely (7)
restaurant again?
nothing can be done) and walks away. How likely are you to say negative
things about the restaurant to other
people? (R)
A.3. Apology scenario How likely are you to recommend
the restaurant to someone who
seeks your advice?
(Continued from control scenario) You call for the server who How likely are you to discourage
approaches you immediately. You complain about the smoking cus- friends or relatives from visiting the
tomers. The severer sincerely apologizes to you but then explains restaurant? (R)
that there is nothing could be done about the customers. Note. (R) indicates that item was reverse coded.
 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192 191

References Huang, W.H., 2010. Other-customer failure: effects of perceived employee effort and
compensation on complainer and non-complainer service evaluations. Journal
of Service Management 21 (2), 191–211.
Allinson, C.N., Hayes, J., 1996. The cognitive style index: a measure of institu-
Huang, W.H., Lin, Y.C., Wen, Y.C., 2010. Attributions and outcomes of customer
tion analysis of organisational research. Journal of Management Studies 33 (1),
misbehavior. Journal of Business and Psychology 25 (1), 51–161.
119–135.
Hui, M.K., Bateson, J.E.G., 1991. Perceived control and the effects of crowding and
Abramson, L.Y., Seligman, M., Teasdale, J.D., 1978. Learned helplessness in humans:
consumer choice on the service experience. Journal of Consumer Research 18
critique and reformulation. Journal of Abnormal Psychology 87 (1), 49–74.
(2), 174–184.
Andreassen, T.W., 2000. Antecedents to satisfaction with service recovery. European
Iacobucci, D., Ostrom, A., 1993. Gender differences in the impact of core and rela-
Journal of Marketing 34 (1/2), 156–175.
tional aspects of services on the evaluation of service encounters. Journal of
Arnould, E.J., Price, L.L., 1993. River magic: extraordinary experience and the
Consumer Psychology 2 (3), 257–286.
extended service encounter. Journal of Consumer Research 20 (1), 24–45.
Jou, J.,Yang, C.M., 2007. Jaycustomer behavior and its effects on target customers
Baker, J., 1987. The role of the environment in the marketing service: the consumer
cognitive and emotive responses. In: Marketing Theory and Applications in 2007
perspective. In: Czepiel, J.A., Congram, C.A., Shanahan, J. (Eds.), The Services Chal-
AMA Winter Educators’ Conference, American Marketing Association, Chicago,
lenge: Integrating for Competitive Advantage. American Marketing Association,
IL, pp. 55–56.
Chicago, IL, pp. 79–84.
Kanousi, A., 2005. An empirical investigation of the role of culture on service recovery
Becker, J., Ayman, R., Korabik, K., 2002. Discrepancies in self/subordinates’
expectations. Managing Service Quality 15 (1), 57–69.
perceptions of leadership behavior: leader’s gender, organizational con-
Lam, S.Y., Shankar, V., Erramilli, M.K., Murthy, B., 2004. Customer value, satisfaction,
text, and leader’s self-monitoring. Group Organization Management 27 (2),
loyalty, and switching costs: an illustration from a business-to-business service
226–244.
context. Journal of the Academy of Marketing Science 32 (3), 293–331.
Bitner, M.J., 1992. Servicescapes: the impact of physical surroundings on customers
Leventhal, G., Garcia, V.L., 1991. An examination of personal and situational factors
and employees. Journal of Marketing 56 (2), 57–71.
which affect female managers and their employees. Psychological Reports 68,
Bitner, M.J., Booms, B.H., Mohr, L.A., 1994. Critical service encounters: the employee’s
835–848.
viewpoint. Journal of Marketing 59 (4), 95–106.
Levesque, T.J., McDougall, G.H.G., 2000. Service problems and recovery strategies:
Bitner, M.J., Booms, B.H., Tetreault, M.S., 1990. The service encounter: diagnosing
an experiment. Canadian Journal of Administrative Sciences 17 (1), 20–37.
favourable and unfavourable incidents. Journal of Marketing 54 (1), 71–84.
Lovelock, C., 1994. Product Plus: How Product + Service = Competitive Advantage.
Blodgett, J.G., Hill, D.J., Tax, S.S., 1997. The effects of distributive, procedural, and
McGraw-Hill, New York.
interactional justice on post complaint behavior. Journal of Retailing 73 (2),
Martin, C.L., 1996. Consumer-to-consumer relationships: satisfaction with other
185–210.
consumers’ public behavior. Journal of Consumer Affair 30 (1), 146–169.
Boo, H.C., Ismal, A.F., Saad, N.H.M., 2008. Effectiveness of service recovery strategy
Martin, C.L., Pranter, C.A., 1989. Compatibility management: customer-to-customer
on consumers’ perceived justice: a multi-cultural perspective in Malaysia. Revue
relationships in service environments. Journal of Services Marketing 3 (3), 5–15.
Tourisme 16, 69–78 (October).
Mattila, A.S., Cho, W., Ro, H., 2009. The joint effects of service failure mode, recovery
Boshoff, C., 1997. An experimental study of service recovery options. International
effort, and gender on customers’ post-recovery satisfaction. Journal of Travel
Journal of Service Industry Management 8 (2), 110–130.
and Tourism Marketing 26 (2), 120–128.
Bougie, R., Pieters, R., Zeelenberg, M., 2003. Angry customers don’t come back, they
Mattila, A.S., Grandey, A.A., Fisk, G.M., 2003. The interplay of gender and affec-
get back: the experience and behavioral implications of anger and dissatisfaction
tive tone in service encounter satisfaction. Journal of Service Research 6 (2),
in services. Journal of the Academy of Marketing Science 31 (4), 377–393.
136–143.
Brew, F.P., Cairns, D.R., 2004. Do culture or situational constraints determine choice
Mattila, A.S., Patterson, P.G., 2004. The impact of culture on consumers’ perceptions
of direct or indirect styles in intercultural workplace conflicts. International
of service recovery efforts. Journal of Retailing 80 (3), 196–206.
Journal of Intercultural Relations 28 (5), 331–352.
Maxham III, J.G., 2001. Service recovery’s influence on consumer satisfaction, posi-
Chiu, C.Y., Tsang, S.C., Yang, C.F., 1988. The role of face situation and attitudinal
tive word-of-mouth, and purchase intentions. Journal of Business Research 54
antecedents in Chinese consumer complaint behavior. The Journal of Social
(1), 11–24.
Psychology 128 (2), 173–180.
McColl-Kennedy, J.R., Daus, C.S., Sparks, B.A., 2003. The role of gender in reactions
Coombs, W.T., 2007. Protecting organization reputations during a crisis: the devel-
to service failure and recovery. Journal of Service Research 6 (1), 66–82.
opment and application of situational crisis communication theory. Corporate
McColl-Kennedy, J.R., Sparks, B.A., 2003. Application of fairness theory to service
Reputation Review 10 (3), 163–176.
failures and service recovery. Journal of Service Research 5 (3), 251–266.
Darley, W.K., Smith, R.E., 1995. Gender differences in information processing strate-
Meyers-Levy, J., 1988. The influence of sex roles on judgment. Journal of Consumer
gies: an empirical test of the selectivity model in advertising response. Journal
Research 14 (4), 522–530.
of Advertising 24 (1), 41–56.
Miao, L., Mattila, A.S., Mount, D., 2011. Other customers in service encounter: a script
Davidow, M., 2000. The bottom line impact of organizational responses to customer
theoretical perspective. International Journal of Hospitality Management 30 (4),
complaints. Journal of Hospitality and Tourism Research 24 (4), 473–490.
933–941.
Elliot, A.J., 2006. The hierarchical model of approach-avoidance motivation. Motiva-
Mohr, L.A., Bitner, M.J., 1995. The role of employee effort in satisfaction with service
tion and Emotion 30 (2), 111–116.
transactions. Journal of Business Research 32 (3), 239–252.
Fischer, E., Arnold, S.J., 1990. More than a labor of love: gender roles and Christmas
Mount, D.J., Mattila, A., 2000. The final opportunity: the effectiveness of a cus-
gift shopping. Journal of Consumer Research 17 (3), 333–345.
tomer relations call center in recovering hotel guests. Journal of Hospitality and
Fisk, R.P., Grove, S.J., Harris, L.C., Keeffe, D.A., Daunt, K.L., Russell-Bennett, R.,
Tourism Research 24 (4), 514–525.
Wirtz, J., 2010. Customers behaving badly: a state of the art review, research
Patterson, P.G., Cowley, E., Prasongsukarn, K., 2006. Service failure recovery: the
agenda and implications for practitioners. Journal of Services Marketing 24 (6),
moderating impact of individual-level culture value orientation on perceptions
417–429.
of justice. International Journal of Research in Marketing 23 (3), 263–277.
Fullerton, R.A., Punj, G., 1993. Choosing to misbehave: a structural model of aberrant
Perdue, B.C., Summers, J.O., 1986. Checking the success of manipulations in market-
consumer behavior. Advances in Consumer Research 20 (1), 570–574.
ing experiments. Journal of Marketing Research 23 (4), 317–326.
Goodwin, C., Ross, I., 1992. Consumer responses to service failure: influence of pro-
Reynolds, K.L., Harris, L.C., 2006. Deviant customer behavior: an exploration of front-
cedural and interactional fairness perceptions. Journal of Business Research 25
line employee tactics. The Journal of Marketing Theory and Practice 14 (2),
(2), 149–163.
95–111.
Grewal, D., Roggeveen, A.L., Tsiros, M., 2008. The effect of compensation on repur-
Reynolds, K.L., Harris, L.C., 2009. Dysfunctional customer behavior severity: an
chase intentions in service recovery. Journal of Retailing 84 (4), 424–434.
empirical examination. Journal of Retailing 85 (3), 321–335.
Grove, S.J., Fisk, R.P., 1997. The impact of other customers on service experiences: a
Roth, S., Cohen, L.J., 1986. Approach, avoidance, and coping with stress. American
critical incident examination of getting along. Journal of Retailing 73 (1), 63–85.
Psychologist 41 (7), 813–819.
Gruber, T., Szmigin, I., Voss, R., 2009. Handling customer complaints effectively:
Rust, R.T., Cooil, B., 1994. Reliability measures for qualitative data: theory and impli-
a comparison of the value maps of female and male complainants. Managing
cations. Journal of Marketing Research 31 (1), 1–14.
Service Quality 19 (6), 636–656.
Schmitt, B.H., Leclerc, F., Dubé-Rioux, L., 1988. Sex typing and consumer behav-
Han, H., Back, K.J., Barrett, B., 2009. Influencing factors on restaurant customers’
ior: a test of gender schema theory. Journal of Consumer Research 15 (1),
revisit intention: the roles of emotions and switching barriers. International
122–128.
Journal of Hospitality Management 28 (4), 563–572.
Smith, A.K., Bolton, R.N., 1998. An experimental investigation of customer reactions
Harris, L.C., Reynolds, K.L., 2003. The consequences of dysfunctional customer behav-
to service failure and recovery encounters – paradox or peril? Journal of Service
ior. Journal of Service Research 6 (2), 144–161.
Research 1 (1), 65–81.
Harris, L.C., Reynolds, K.L., 2004. Jaycustomer behavior: an exploration of types
Smith, A.K., Bolton, R.N., Wagner, J., 1999. A model of customer satisfaction with
and motives in the hospitality industry. Journal of Services Marketing 18 (5),
service encounters involving failure and recovery. Journal of Marketing Research
339–357.
36 (3), 356–372.
Hess Jr., R.L., Ganesan, S., Klein, N.M., 2003. Service failure and recovery: the impact
Tabachnik, B.G., Fidell, L.S., 1996. Using Multivariate Statistics. Harper Collins Col-
of relationship factors on customer satisfaction. Journal of the Academy of Mar-
lege, New York.
keting Science 31 (2), 127–145.
Tax, S.S., Brown, S.W., Chandrashekaran, M., 1998. Customer evaluations of service
Heung, V.C.S., Lam, T., 2003. Customer complaint behavior towards hotel restaurant
complaint experiences: implications for relationship marketing. Journal of Mar-
services. International Journal of Contemporary Hospitality Management 15 (5),
keting 62 (2), 60–76.
283–289.
Ting-Toomey, S., Gao, G., Trubisky, P., Yang, Z., Kim, H.S., Lin, S.L., Nishida, T.,
Huang, W.H., 2008. The impact of other-customer failure on service satisfaction.
1991. Culture, face maintenance, and styles of handling interpersonal conflict: a
International Journal of Service Industry Management 19 (4), 521–536.
192 H.C. Boo et al. / International Journal of Hospitality Management 35 (2013) 180–192

study in five cultures. International Journal of Conflict Management 2 (4), 275– Wirtz, J., Chew, P., 2002. The effects of incentives, deal proneness, satisfaction and tie
296. strength on word-of-mouth behavior. International Journal of Service Industry
Voorhees, C.M., Brady, M.K., Horowitz, D.M., 2006. A voice from the silent masses: Management 13 (2), 141–162.
an exploratory and comparative analysis of noncomplainers. Journal of the Wirtz, J., Mattila, A.S., 2004. Consumer responses to compensation, speed of recov-
Academy of Marketing Science 34 (4), 513–527. ery and apology after a service failure. International Journal of Service Industry
Walster, E., Berscheid, E., Walster, G.W., 1973. New directions in equity research. Management 15 (2), 50–166.
Journal of Personality and Social Psychology 25 (2), 151–176. Wong, N.Y., 2004. The role of culture in the perception of service recovery. Journal
Weiner, B., 2000. Attributional thoughts about consumer behavior. Journal of Con- of Business Research 57 (9), 957–963.
sumer Research 27 (3), 382–387.