Why Service Recovery Fails? Examining the Roles of Restaurant Type and Failure Severity in Double Deviation With Justice Theory

Abstract

Due to the increasingly competitive nature of the industry, the prevalence of service failure in restaurants has made a satisfactory service recovery critical for retaining customers. However, the success rate of service recovery is far from satisfactory. Informed by Rawls’s justice theory, this study explored service recovery failures (double service failure) in a restaurant setting. Results from our experiment indicate that the effects of different types of service recovery failure on postrecovery evaluations vary across two situational factors: restaurant type and failure severity. Specifically, procedural injustice (low-resolution speed) was found to exert more influence on word-of-mouth intentions in a quick-service restaurant than a full-service restaurant. For failures of high severity (vs. low severity), distributive injustice (no compensation offered) is found to be more impactful. Theoretical and managerial implications are discussed.

Keywords

justice theory service recovery failure double deviation failure severity restaurant types

Introduction

Regardless of the industry or how rigorously the service is performed, service errors are inevitable (Park et al., 2014). Studies have found that service errors can significantly contribute to customer defections (Nikbin et al., 2016) and revenue loss (De Matos et al., 2013). It is reported that 33% of U.S. consumers would consider switching service providers after one single experience of poor service (American Express, 2017). As a result, an estimated US$1.6 trillion was lost from customer switching due to poor services (Accenture, 2016). Furthermore, it has been found that American consumers share their negative service experience with an average of 15 people (American Express, 2017), and the most viewed, shared, and trusted reviews are those that describe a negative service experience (Shin et al., 2018). Thus, the impacts of service failure are more likely to snowball due to the increasing importance of social media on consumer perceptions and purchase intentions. With the ever-increasing use of social media as an opportunity to review services, service failure or recovery failures can be instantly “seen” and “heard” by a mass of potential restaurant patrons.

As such, service failure was found to have a significant negative impact on a service firm’s profitability (Chuang et al., 2012), employee morale (Hur & Jang, 2016), and brand image (Hazée et al., 2017; La & Choi, 2012). These impacts of poor customer experience caused by service failure or failed service recovery attempts signify the need for more successful service recovery strategies. Unfortunately, despite firms’ efforts to demonstrate their dedication to service quality, the success rate of “service recoveries”—the process by which they respond to a service failure—has yet to reach satisfactory levels (Van Vaerenbergh & Orsingher, 2016).

Although service failures have been reported to plague all types of services (Van Vaerenbergh & Orsingher, 2016), the impacts of service failures have been suggested to be context-specific. Compared with other service types, it has been found that hospitality-oriented services (including restaurants) undergo intense scrutiny for their service recovery performance (Swanson & Hsu, 2011). Aspects of service recovery attempts are magnified because customers who have experienced a service failure generally become more “aware” and emotionally involved in the service experience (Cai & Qu, 2018). For example, once a restaurant customer found something wrong with his or her order, the customer is likely to be more mindful of the service steps to follow. These service steps could include the speed in which the issue is resolved or the disposition of the service providers handling the recovery. It has been reported that customers who encounter a failed service recovery attempt have experienced a double deviation (Bitner et al., 1990; Joireman et al., 2013). Although double deviations have been found to severely aggravate customers and damage an organization’s reputation (Basso & Pizzutti, 2016), previous studies that examined double deviation and incorporated Rawls’s (1971) justice theory lack specificity with regard to factors that may or may not influence service recovery. Thus, they are limited to an elementary and potentially underdeveloped understanding of the relationship between service recovery and justice.

This study argues that the success of a service recovery depends, in part, on specific contextual factors, namely, restaurant types and failure severity in this study. Thus, the goal of this study is to better understand how Rawls’s (1971) justice theory can be best utilized in understanding service recovery failure. Specifically, this study examined how the effects of different recovery failures (i.e., distributive injustice, procedural injustice, and interactional injustice) on postrecovery evaluations vary across restaurant type (i.e., quick-service restaurant vs. full-service restaurant) and initial service failure severity (high severity vs. low severity).

Literature Review

Justice Theory

Although several theoretical frameworks have been used to examine service recovery (Wen & Chi, 2013), the most popular theory is Rawls’s (1971) justice theory. First introduced in the services marketing literature by Tax et al. (1998), justice theory is a powerful predictor of postrecovery satisfaction (Kim et al., 2012). Justice theory is based on the notion that the recovery experience is evaluated in terms of fairness or justice. Based on justice theory (Rawls, 1971), service marketing researchers have commonly conceptualized recovery justice as being comprised of three interrelated dimensions: distributive justice, procedural justice, and interactional justice. According to justice theory, postrecovery satisfaction is significantly influenced by a customer’s perception of justice demonstrated during the service recovery (Gelbrich & Roschk, 2011).

Justice theory is most commonly described as being comprised of three dimensions: distributive justice, procedural justice, and interactional justice (Rawls, 1971). Described as efforts of provider atonement, distributive justice is characterized by both tangible and intangible forms of compensation (Kim et al., 2009). Common examples of distributive justice provided by services marketing researchers include discounts, refunds, replacement items, and coupons (Park et al., 2014). Procedural justice has been described as the customers’ evaluations of the policies, procedures, and methods of firms used to resolve a conflict (Maxham & Netemeyer, 2002). With regard to restaurant service recovery, previous research has focused on the promptness of fielding complaints (Van Vaerenbergh & Orsingher, 2016), as well as the service provider’s ability to demonstrate flexibility and promptness in problem-solving (Nikbin & Hyun, 2015). Interactional justice is associated with the interactional aspects, as opposed to the formal procedures/policies or the outcomes associated with the service recovery (Swanson & Hsu, 2011). This perception is presumed to be demonstrated through the physical and verbal cues provided by the service provider(s) (Wen & Chi, 2013). To this point, Siu et al. (2013) incorporated the following indicators in evaluating interactional justice during service recovery: honesty, courtesy, fairness, and ethical behavior.

It has been suggested that justice theory is particularly useful when studying restaurants, as justice theory acknowledges the instrumental as well as the relational aspects of service transactions (Nikbin et al., 2016). Subsequently, restaurants have been a prominent setting among marketing services researchers examining service recovery, as restaurant customers have been suggested to experience service failures more often than any other service opportunity (Yoo et al., 2006).

Two of the most common postrecovery evaluations examined in previous research incorporating justice theory are postrecovery satisfaction and word of mouth (WOM; Migacz et al., 2018). Several theoretical approaches have been used to examine postrecovery satisfaction, including, but not limited to, expectancy disconfirmation theory (Nguyen et al., 2012), equity theory (Piaralal et al., 2016), and mental accounting theory (Chuang et al., 2012). However, it has been suggested that justice-based satisfaction had higher predictive power than other measures of customer satisfaction following a service recovery failure (Oliver, 1997) and that justice theory provides for specific and actionable implications for hospitality managers (Kim et al., 2009).

Services marketing researchers have also suggested that WOM communications are critical outcomes of service recovery (Choi & Choi, 2014) due to their influence on customer expectations and future purchase probability (Vázquez-Casielles et al., 2013). It has been suggested that hospitality firms, in particular, are pressured to demonstrate service excellence (Su et al., 2017), as potential restaurant patrons have been found to proactively search for information “high in credence and experience qualities” (Swanson & Hsu, 2011, p. 514). Also, advances in (and access to) technology have helped to create additional channels for WOM, including emails, blogs, online discussion forums, and product review websites (Ring et al., 2016). For example, Ong (2012) examined how respondents utilized online customer reviews in choosing a restaurant. Results indicated that respondents use restaurant reviews differently and would search for more reviews depending on the restaurant location, the costs associated with the restaurant, and the restaurant type.

Service Recovery Failure: A Double-Deviation Situation

Bitner et al. (1990, p. 80) defined a double deviation as a “perceived inappropriate and/or inadequate response to failures in the service delivery system” (i.e., an unsuccessful attempt to restore customer’s satisfaction after the initial service failure). Double deviation was found to intensify customers’ dissatisfaction and negative reactions toward the initial service failure (Basso & Pizzutti, 2016), which further lead to complaints and switch service providers (Casado-Díaz et al., 2008). Given that failed recovery is the leading cause of switching behavior (Haj-Salem & Chebat, 2013) and the subsequent damage that a double deviation could bring to a company’s reputation and profitability, it is imperative to understand what causes a failed service recovery.

Research attention to double deviation is surprisingly scarce (Basso & Pizzutti, 2016; Casado-Díaz et al., 2007, 2008). Past research on service recovery has been predominantly focused on understanding the initial service failure and developing a recipe for an effective recovery. Still, little attention has been paid to the occurrence of unsatisfactory service recovery (Basso & Pizzutti, 2016). Among the limited studies, a few aimed at formulating the recovery strategy from double deviation but with mixed results. For instance, Joireman et al. (2013) showed that compensation (i.e., distributive justice) and apology (i.e., interactional justice) were most effective, whereas Basso and Pizzutti (2016) found that financial compensation was less adequate.

In light of the above mixed findings in double deviation, it is clear that there is no one-size-fits-all approach. We argue that these limited and contradictory findings may be a result of an oversight of situational factors such as the service level (e.g., restaurant type) and initial failure severity. Customers have clear expectations of service recovery (Michel et al., 2009): These expectations are likely shaped by the context (Park et al., 2014), which in turn would contribute to moderating perceptions of the service recovery process (Ha & Jang, 2009; Sparks & Fredline, 2007; Swanson & Hsu, 2011). The following section will discuss the impacts of situational factors (i.e., restaurant type and initial failure severity) on service recovery.

Restaurant Type and Service Recovery

Several studies have previously incorporated specific categories to further understand service recovery strategy, including service failure type (Hoffman et al., 1995), industry type (Mattila, 2001), and country of origin (Dutta et al., 2007). Some hospitality-centric studies have also incorporated a specific restaurant setting (i.e., fast-food, casual, and fine-dining restaurants) as the focus of service recovery research (Babin et al., 2005; Namkung & Jang, 2010). However, no study has previously incorporated multiple restaurant settings and justice theory in service failure research. Therefore, to further increase validity and improve the usability of potential findings, this study has incorporated justice theory in both casual and fast-food dining contexts.

Industry classifications are often based on specific features shared by similar entities (Hrazdil et al., 2013). For restaurants, those descriptors would likely include price levels and the level of food/service quality provided. For example, the National Restaurant Association (NRA) has reported five major restaurant industry segments: quick-service (or fast-food), fast-casual, midscale, moderate (or casual), and fine dining (or upscale). However, overlapping or ambiguous classifications have been an issue in previous research examining restaurant classifications (Canziani et al., 2016). Thus, the current study has limited the scope of the restaurant classifications to “full-service restaurant” and “quick-service restaurant,” as restaurant patrons have recognized these segments as distinct restaurant types (Ponnam & Balaji, 2014).

As previous research suggests that postrecovery evaluations are predicated on customer expectations (Kim et al., 2009), which are context-specific (Mattila, 2001), those expectations are shaped by the environment in which the initial service failure occurred. Customers’ expectations and importance of restaurant attributes vary across restaurant types (Namkung & Jang, 2010). The friendliness of the servers (i.e., interactional justice) is vital for customers in fine-dining restaurants, whereas the speed and cost of the food are critical in a causal/quick-service restaurant setting (Migacz, 2018; Nikbin et al., 2016). Such expectations and attribute importance are likely to transfer to postrecovery evaluations. Therefore, the roles of the three types of justice in service recovery may differ by restaurant type.

Notably, it is argued that compared with a full-service restaurant, procedural and distributive justice following a service failure is more important in a quick-service restaurant, and interactional justice is less critical for quick-service restaurant patrons than for full-service restaurant customers. Noone et al. (2007) stated that customers had clear pacing expectations for different restaurants as they found customers were less satisfied when being rushed in a full-service restaurant than in a quick-service restaurant. Thus, procedural injustice (i.e., the slow speed in which the problem is resolved) is postulated to be more detrimental to a quick-service restaurant setting than a full-service restaurant setting. Furthermore, Mason et al. (2016) found that quick-service restaurant customers’ satisfaction depended more on the perception of value than other restaurant attributes, such as the atmosphere. Thus, it can be suggested that distributive injustice (i.e., lack of compensation) likely matters more for quick-service patrons than full-service customers. Finally, it was found that in a fine-dining restaurant setting, employee’s customer orientation has the most substantial influence on relationship quality (i.e., customer trust and satisfaction; Kim et al., 2006). Combined, these findings illustrate the crucial role of interactive justice in full-service dining as opposed to quick-service settings.

In light of the above discussion, the following hypotheses are proposed:

Hypothesis 1 (H1): The type of restaurant (i.e., quick-service restaurant vs. full-service restaurant) will moderate the effects of the three types of injustice on satisfaction.

Hypothesis 1a (H1a): Distributive injustice will have a stronger impact on satisfaction in a quick-service restaurant setting (vs. a full-service restaurant setting).

Hypothesis 1b (H1b): Interactional injustice will have a stronger impact on satisfaction in a full-service restaurant setting (vs. a quick-service restaurant setting).

Hypothesis 1c (H1c): Procedural injustice will have a stronger impact on satisfaction in a quick-service restaurant setting (vs. a full-service restaurant setting).

Hypothesis 2 (H2): The type of restaurant (i.e., quick-service restaurant vs. full-service restaurant) will moderate the effects of the three types of injustice on WOM intentions.

Hypothesis 2a (H2a): Distributive injustice will have a stronger impact on WOM intentions in a quick-service restaurant setting (vs. a full-service restaurant setting).

Hypothesis 2b (H2b): Interactional injustice will have a stronger impact on WOM intentions in a full-service restaurant setting (vs. a quick-service restaurant setting).

Hypothesis 2c (H2c): Procedural injustice will have a stronger impact on WOM intentions in a quick-service restaurant setting (vs. a full-service restaurant setting).

Failure Severity and Service Recovery

Service failure severity is defined as “a customer’s perceived intensity of a service problem” (Weun et al., 2004, p. 135). A customer’s evaluation of a recovery attempt may be influenced by the perceived magnitude of the service error (Susskind & Viccari, 2011). Several researchers have suggested that the higher the magnitude of the service failure, the greater the dissatisfaction associated with the initial service transaction, and thus the greater the challenge for the service provider to enact a successful service recovery (Hur & Jang, 2016; Magnini et al., 2007).

Moreover, the magnitude of a service failure also influences the type of recovery attempts necessary to regain customer’s satisfaction (Betts et al., 2011). For instance, a sincere apology is likely to be sufficient to recover a restaurant customer’s 10-min wait for being seated, but it might not be able to adequately address customers’ dissatisfaction due to a foreign object found in food. According to De Matos et al. (2007), the service recovery paradox only existed in a modest failure severity situation in which a superior recovery was performed. Subsequently, the magnitude of the service failure should be a consideration when service recovery attempts are appropriated (Gelbrich & Roschk, 2011).

Given the critical role of severity in a successful service recovery (Swanson & Hsu, 2011), it is surprising that most previous service recovery studies held failure severity level constant (La & Choi, 2019). Among the limited number of studies focused on failure severity, Weun et al. (2004) and Mattila (2001) investigated the relationship between failure severity and justice perception. Weun et al. (2004) found a moderating effect of failure severity on the relationship between distributive justice and satisfaction, but no moderation was found on the interactional justice/satisfaction relationship. However, Weun et al. (2004) only focused on two of the three justice types (distributive justice and interactional justice). While Mattila (2001) found a consistent result for distributive justice, her study also showed a significant effect of failure severity on interactional justice, which is inconsistent with Weun et al. (2004). In addition, Mattila (2001) did not find a significant effect of failure severity on procedural justice.

This study intends to extend the service recovery literature by examining service failure severity in relation to all three types of justice (Chuang et al., 2012; Mattila, 2001) and, subsequently, provide viable recovery strategies for the restaurant industry. Based on Weun et al.’s (2004) and Mattila’s (2001) findings, it is probable that distributive justice can result in better postrecovery evaluations when failure severity is low, relative to when failure severity is high. Following Mattila (2001), it is postulated that procedural justice might not vary across severity levels. The relatively contradictory results on interactional justice warrant an additional investigation. Therefore, the following hypotheses are proposed:

Hypothesis 3 (H3): The severity of a service failure (high vs. low severity) will moderate the effects of the three types of justice on satisfaction.

Hypothesis 3a (H3a): Distributive injustice will have a stronger impact on satisfaction when service failure severity is high (vs. low failure severity).

Hypothesis 3b (H3b): Interactional injustice will have a stronger impact on satisfaction when service failure severity is high (vs. low failure severity).

Hypothesis 3c (H3c): Procedural injustice will have a stronger impact on satisfaction when service failure severity is high (vs. low failure severity).

Hypothesis 4 (H4): The severity of a service failure (high vs. low severity) will moderate the effects of the three types of justice on WOM intentions.

Hypothesis 4a (H4a): Distributive injustice will have a stronger impact on WOM intentions when service failure severity is high (vs. low failure severity).

Hypothesis 4b (H4b): Interactional injustice will have a stronger impact on WOM intentions when service failure severity is high (vs. low failure severity).

Hypothesis 4c (H4c): Procedural injustice will have a stronger impact on WOM intentions when service failure severity is high (vs. low failure severity).

Method

Research Design

To test the proposed hypotheses, this study used a 4 (justice: baseline vs. procedural injustice vs. interactional injustice vs. distributive injustice) × 2 (restaurant types: quick-service vs. full-service) × 2 (severity: high vs. low) between-subject experiment. Participants were randomly assigned to one of the 16 scenarios involving a dining experience (see the appendix for sample experimental scenario).

The four justice conditions were developed by manipulating three components: the speed in which the problem was resolved (procedural justice), the presence/absence of the apology from the server (interactional justice), and the presence/absence of the monetary compensation (distributive justice). In the procedural injustice condition, the server took more than 30 min to solve the problem, whereas in the other justice conditions, procedural justice was demonstrated by providing the customer with a solution to the service error within 10 min of the occurrence. In the interactional injustice condition, the server did not apologize, whereas in the other justice conditions, the server provided an apology. In the distributive injustice condition, the bill of the meal was taken off as monetary compensation for the service failure, whereas in the other justice conditions, no financial compensation was offered. An error-free service recovery scenario, a scenario in which all justice dimensions were demonstrated, served as the baseline condition.

Restaurant types were manipulated by providing participants one of the two scenarios in which they found themselves dining at a fast-food restaurant (quick-service condition) or a full-service restaurant (full-service restaurant condition) at the beginning of the scenario. Service failure severity was manipulated using two types of service errors. In the high severity condition, the service failure was identified as an undesirable object (a piece of glass in the dish). In the low severity condition, the service failure was identified as being provided the wrong meal.

In the survey’s introduction, participants were asked to provide information about their average dining experience, including the frequency of dining out. Specifically, participants were asked to volunteer information regarding previous experiences of service recovery involving restaurants; this information was then used as the basis for additional questions focused on the participant’s best and worst experience utilizing the critical incident technique (CIT). This was followed by scenario scripts in which participants were asked to imagine that they were dining in a restaurant of their choice with a group of friends, in which the dining experience was interrupted by a service error. Each participant was randomly assigned to one of the 16 conditions. After reading the script, participants were then asked a series of manipulation check questions and a series of questions related to satisfaction and WOM intentions. Finally, the survey finished with demographic questions.

Participants and Data Collection

Participants were recruited through the internet using Amazon’s Mechanical Turk (MTurk). MTurk is an internet-based human intelligence marketplace with approximately 500,000 individuals, referred to as “workers,” who voluntarily complete tasks in return for a monetary payment (Amazon’s Mechanical Turk, 2014). There are at least three advantages of using MTurk for sampling purposes: First, the size of the sample pool is mostly larger than university’s sample pools; second, the demographic background of participants is more diverse than college and online samples; and third, previous studies using MTurk have found that the quality of the data obtained from MTurk had the same, if not better, reliability as that from conventional sampling methods (Byun & Jang, 2015; Mason & Suri, 2012).

To ensure the quality of the data, workers who participated in this survey were required to be located in the United States and have a “master” qualification granted by MTurk. “Masters” are elite groups of “Workers” who have demonstrated accuracy on specific types of human intelligence tasks (HITs) on MTurk. “Workers” achieve a “Masters” distinction by consistently completing HITs of a certain type with a high degree of accuracy across a variety of requesters, and “Masters” must continue to pass MTurk’s statistical monitoring to remain qualified (Amazon’s Mechanical Turk, 2014).

A total of 594 participants completed the survey in exchange for a small payment. As shown in Table 1, the average age of the sample was 37.6 years. About 64.5% of the participants were female and about 47.3% held bachelor’s degrees. Approximately 59.1% of the participants had household incomes of US$40,000 or above. More than 95% of the participants dined out at least once per week, and about 50% of them acknowledged having work experience in the restaurant industry.

Table 1.

Profile of Respondents.

	Frequency	%	M	SD
Gender
Male	211	35.5
Female	383	64.5
Age			37.6	12.93
18–29	201	33.9
30–39	185	31.1
40–49	81	13.6
50–59	79	13.3
60+	47	7.9
Missing	1	0.2
Years of education
Less than high school	3	0.5
High school graduates	70	11.8
Some college	139	23.4
2-year college	74	12.5
4-year college	209	35.2
Professional degree	27	4.5
Graduate school	72	12.1
Annual household income
Under US$20,000	101	17
US$20,000–US$39,999	142	23.9
US$40,000–US$59,999	125	21
US$60,000–US$79,999	93	15.6
US$80,000–US$99,999	53	8.9
US$100,000–US$150,000	60	10.1
Over US$150,000	20	3.4
On average, how many meals per week do you eat out?
0	27	4.5
1	162	27.3
2	138	23.2
3	101	17
4	57	9.6
5	49	8.2
More than 6	60	10.2
Have you ever worked in a restaurant?
Yes	295	49.7
No	299	50.3
Have you ever, while eating your meal in a restaurant, found something in your plate that did not belong (e.g., hair, glass)?
Yes	362	60.9
No	232	39.1
Have you ever, while dining in a restaurant, been served an order that was not correct?
Yes	483	81.3
No	111	18.7

Measurements

The questionnaire was composed of three sections. In the first section, participants were asked about their experience with service failure and service recovery in a restaurant using CIT open-ended questions. The second section included the assigned scenario script and a series of manipulation check questions. The third section assessed participants’ satisfaction and WOM intentions. Satisfaction with service recovery was measured with four items adapted from Wen and Chi (2013): “The restaurant provided a service recovery that met my needs,” “I am satisfied with the restaurant handling of this particular problem,” “In my opinion, the restaurant provided a satisfactory resolution,” and “Regarding this particular event, I am satisfied with the restaurant” (Cronbach’s α = .96).

Intentions to spread WOM was assessed with the scale adopted from Wen and Chi (2013): “I will recommend this restaurant to my friends and relatives,” “I will provide positive comments about this restaurant on social media sites,” and “I will encourage my friends and relatives to choose this restaurant” (Cronbach’s α = .95). All items in the third section used 7-point Likert-type scales anchored by 1 = strongly disagree and 7 = strongly agree. The survey was finished with demographic questions such as gender, age, education, income, and whether they have ever worked in a restaurant. In terms of restaurant work experience, participants were asked, “Have you ever worked in a restaurant?” (yes/no).

Results

Service Recovery Experience

Among the 594 participants, 60.9% (n = 362) of them have found unintentional items on their plates when dining out. Of those participants who have experienced accidental things on their plates, only 67.7% of them received a response from the restaurant. In addition, 81.3% of participants (n = 483) reported being served a meal that they did not order. Among these participants, this type of service failure happened averagely 15.15% of the time they dined in restaurants. These findings indicate that the two service failure scenarios (foreign objects in one’s plate and being served an incorrect order) frequently occur in the restaurant industry. Thus, the inclusion of these two service failures was deemed acceptable representations of common service failures in the restaurant industry.

Manipulation Checks

The manipulation for restaurant type was checked by a binary question: “Based on the story you have just read, what type of restaurant was described?” To assess justice manipulations, three binary questions were asked: “Based on the story you have just read, how long did you have to wait to have your problem resolved?” “Based on the story you have just read, did the server apologize?” and “Did you receive any compensation?” Only participants who got all binary manipulation check questions were included in the data analysis. The manipulation for severity was checked with an 11-point Likert-type scale question: “Based on the story you have just read, do you feel the service mistake in the story was severe (0 = not at all severe, 10 = extremely severe)?” Results showed that the perceived severity was significantly higher in the high severity group than the low severity group (M_{low_severity} = 4.17 vs. M_{high_severity} = 7.4, t = 16.63, p = .00).

Hypothesis Testing

The items for each dependent variable (i.e., satisfaction and intentions to spread WOM) were averaged. Two 4 (justice: baseline vs. procedural injustice vs. interactional injustice vs. distributive injustice) × 2 (restaurant type: quick-service vs. full-service) × 2 (severity: high vs. low) analysis of variance (ANOVA) tests were performed on satisfaction and WOM, respectively, and no significant three-way interaction was found for both dependent variables. As a three-way interaction effect was not theorized and the interaction was not significant, the data were randomly split into two sub-data sets to test the interaction effects between injustice and restaurant type, and between injustice and severity.

Injustice was recoded into three dummy variables: distributive injustice (yes = 1, no = 0), interactional injustice (yes = 1, no = 0), and procedural injustice (yes = 1, no = 0) to allow a more focused regression analysis. Similarly, restaurant type (full-service restaurant = 1, quick-service restaurant = 0) and severity (high severity = 1, low severity = 0) were also coded as dummy variables.

The effects of restaurant types and the three types of injustice

Satisfaction

A multiple regression analysis was performed to investigate the interaction effects between the three types of injustice and restaurant types on postrecovery satisfaction. Table 2 demonstrates the results and showed that there was no significant interaction effect on satisfaction. Thus, H1 (i.e., H1a, H1b, and H1c) was not supported.

Table 2.

Coefficients (and Standard Errors) From Regression Analyses Predicting Satisfaction and WOM.

Source of Variance	Model 1Satisfaction	Model 2WOM
Constant	6.56*** (.40)	5.86*** (.49)
Distributive injustice	−1.92** (.55)	−2.67*** (.67)
Interactional injustice	−1.42** (.52)	−2.74*** (.64)
Procedural injustice	−1.85** (.56)	−2.96*** (.68)
Restaurant type	−0.10 (.48)	−0.53 (.59)
Distributive Injustice × Restaurant Type	−0.03 (.68)	1.12 (.82)
Interactional Injustice × Restaurant Type	0.24 (.66)	1.19 (.80)
Procedural Injustice × Restaurant Type	1.05 (.69)	2.34** (.83)
R ²	.22	.25

Note. WOM = word of mouth.

p< .05. **p < .01. ***p < .001.

As expected, significant main effects of the three injustice were found. Compared with the baseline condition (i.e., the reference group), all three injustice types significantly decreased satisfaction (b_distributive = −1.92, t = −3.47, p = .001; b_{interactional} = −1.42, t = −2.72, p = .007; b_procedural = −1.85, t = −3.29, p = .001).

WOM

A multiple regression analysis was performed to investigate the interaction effects between the three types of injustice and restaurant types on WOM. Table 2 indicates a significant interaction effect between restaurant type and procedural injustice (b_{Type × Procedural} = 2.34, t = 2.82, p = .006). Specifically, when it took the restaurant 30 min to resolve the service failure (procedural injustice), intentions to spread WOM were lower in the quick-service restaurant condition than in the full-service restaurant condition by 1.81 on a 7-point scale. Therefore, H2c was supported by the data, but H2a and H2b were not.

The effects of failure severity and the three types of justice

Satisfaction

Another half of the data set was used to perform a multiple regression analysis to investigate the interaction effects between the three types of injustice and failure severity on postrecovery satisfaction. Table 3 demonstrates the results and shows no significant interactions between severity and procedural/interactional injustice, but a significant interaction effect between severity and distributive injustice was found (b_{Severity × Distributive} = −1.83, t = −2.73, p = .007). Specifically, when compensation was not offered (i.e., distributive injustice), satisfaction was significantly lower in the high severity condition than in the low severity condition by 2.48 on a 7-point scale. Thus, H3a was supported by the data, but H3b and H3c were not supported.

Table 3.

Coefficients (and Standard Errors) From Regression Analyses Predicting Satisfaction and WOM.

Source of Variance	Model 1Satisfaction	Model 2WOM
Constant	6.66*** (.29)	6.11*** (.33)
Distributive injustice	−1.80*** (.41)	−2.11*** (.46)
Interactional injustice	−1.61*** (.41)	−2.23*** (.47)
Procedural injustice	−0.66 (.46)	−1.23* (.53)
Severity	−0.65 (.42)	−2.11*** (.46)
Distributive Injustice × Severity	−1.83** (.67)	−0.12 (.74)
Interactional Injustice × Severity	0.03 (.60)	0.55 (.66)
Procedural Injustice × Severity	−0.41 (.62)	0.48 (.70)
R ²	.36	.39

Note. WOM = word of mouth.

p < .05. **p < .01. ***p < .001.

WOM

Table 3 shows the multiple regression analysis results and found no significant interaction effects between severity and injustice types. Thus, H4 was not supported. However, a significant main effect of failure severity (b_severity = −2.11, t = −4.59, p < .001) and significant main effects of the three injustice were found. Compared with the baseline condition (i.e., the reference group), all three injustice types significantly decreased WOM intentions (b_distributive = −2.11, t = −4.64, p < .001; b_{interactional} = −2.23, t = −4.8, p < .001; b_procedural = −1.23, t = −2.34, p = .02). Therefore, H4 (i.e., H4a, H4b, and H4c) was not supported.

Discussions and Implications

The study aimed to explore service recovery failure with Rawls’s (1971) justice theory. Specifically, the purpose of this study was to examine the moderating impact of restaurant type and failure severity on failed service recovery attempts. It was found that the effects of justice types on postrecovery WOM intentions are moderated by restaurant type. Specifically, customers of a quick-service restaurant (vs. a full-service restaurant) are less likely to spread positive WOM when procedural justice is not fulfilled in service recovery. In contrast, no differences were found in the other two justices. In other words, patrons of quick-service restaurants who have experienced a service recovery failure and perceive a lack of promptness in the service recovery attempt are less likely to provide positive WOM. Besides, it was found that when the service failure is perceived to be of high severity (vs. low severity), distributive justice was found to be the most impactful in a service recovery failure. When compensation was not offered, satisfaction was significantly lower in the high severity condition than in the low severity condition.

The present study provides several theoretical contributions to the service recovery literature. The finding that the baseline condition resulted in the highest recovery effectiveness offers further support for the three-dimensional conceptualization of recovery justice introduced by Tax et al. (1998). As service recovery scholars have yet to reach a consensus as to which justice dimension among the three is most critical to performing a successful service recovery (Migacz et al., 2018), the findings of this study further contribute to that goal indirectly by examining justice dimensions via failed recovery attempts. Previous service recovery research has found several potential moderators, such as service type (Mattila, 2001), customer’s relationship quality levels (Ha & Jang, 2009), and customers’ cultural value orientation (Patterson et al., 2006), yet the current study found that the recovery effectiveness of the three dimensions of justice varies across different restaurant types and service failure severity. In other words, the findings of this study are consistent with the notion that service recovery ineffectiveness is context specific (Levesque & McDougall, 2000; Mattila, 2001).

The limited previous research aimed to gain a better understanding of the effect of service failure on recovery efforts has mainly focused on the linear relationship between the magnitude of the service error and customers’ postrecovery evaluation (e.g., Mattila, 1999, 2001; Swanson & Hsu, 2011). It has been previously established that recovery tends to be more difficult when a service failure is considered severe (Mattila, 1999; Smith & Bolton, 1998). Yet, little is known about how to best determine an appropriate recovery strategy given the severity of the service failure. Consistent with Weun et al. (2004), our research shows that severity level moderated the relationship of distributive injustice and satisfaction, but such moderation was not found in the procedural injustice/satisfaction and interactional injustice/satisfaction relationships. When failure severity is high, customers expect restaurants to go the extra mile to make it right. As Migacz et al. (2018) found that service providers’ recovery practices mostly focus on interactional and procedural justice, it is probable that distributive-centered recovery is more difficult for service firms to provide and that the absence of compensation signals provider’s unwillingness to go the extra mile, which is more upsetting for customers who encounter a severe failure.

Nevertheless, no moderation effect was found in the justice/WOM intentions relationships, suggesting that the effects of the four injustice types on WOM intentions were similar between high and low severity conditions. However, Jones et al. (2002) identified specific sociodemographic and personality factors of restaurant patrons that contributed to the dissemination of negative WOM. More recently, Salem et al. (2017) found both age and education levels to have a significant moderating effect on the restaurant patrons’ satisfaction and WOM behavior. Thus, this study’s findings could suggest that proclivity for WOM (positive or negative) is attributable more to the individual restaurant patron’s sociodemographic makeup and personality and less attributable to the perceived severity of the service failure and subsequent service recovery assessment.

Finally, we examined an underexplored yet significant moderator of service recovery effectiveness: restaurant type. Based on an extensive literature review, this study is among the first to investigate the moderating effect of restaurant type in service recovery failure. Previous empirical evidence suggests that customers’ expectations of service recovery in the restaurant industry are different from other service industries (Mattila, 2001). This is due in part to the very nature of food service, as the act of eating “involves an extremely intimate exchange between the environment and the self” (Rozin et al., 1997, p. 68). This study extends this line of research by showing that customers’ expectations of service recovery are also different across different restaurant types.

Moreover, this study found that while the three types of justice dimensions were equally important for customers’ postrecovery satisfaction, procedural justice was found to be of particular importance for fast-food patrons with regard to WOM intentions. These findings are not consistent with Mattila (2001), who found distributive justice to be the most influential and procedural justice to be the least influential justice dimension on WOM intentions. However, it is worth mentioning that Mattila (2001) examined a combination of service recovery options (an apology combined with a 20% discount vs. neither an apology nor compensation) in a restaurant’s long-waiting service failure context. The lack of procedural justice in both the recovery manipulation and the service failure scenario likely contributed to the finding that procedural justice had the least influence on customers’ postrecovery evaluations in the restaurant setting. Although this study examined the effects of individual types of justice on postrecovery evaluations, this contradiction indicates that there is a critical need to better understand the influence of restaurant type on recovery efforts.

This study also offers some practical implications. Based on the results, the baseline conditions (i.e., an error-free recovery) resulted in the highest postrecovery satisfaction and WOM intentions. This indicates that service recovery is more likely to be perceived most fairly when a restaurant’s recovery effort satisfies all three justice dimensions. Subsequently, the most logical managerial recommendation would be for restaurants to consider all three dimensions of justice when designing service recovery policies and strike for an error-free service recovery through continuous employee training and job assessments that appropriately prioritize service recovery.

However, the high frequency of service recovery failures reported in this study suggests an expectation of error-free service recovery in the restaurant industry be unrealistic. What is more, the findings of this study indicate that a one-size-fit-all strategy for service recovery is neither efficient nor effective in retaining customers. Specifically, it is recommended that when customers consider a service mistake as severe, recovery should prioritize compensation and the speed of the resolution. More importantly, restaurant management should make sure customers perceive the compensation and the speed of the response in the service recovery attempt as fair and just.

In addition, this study also found that customers are particularly unforgiving with severe service breakdowns compared with minor service mistakes as an error-free recovery results in a lower level of satisfaction and WOM intentions in the high severity condition than in the low severity condition. This implies that the prevention of severe failures might be more important than recovery. Therefore, restaurant managers should first better understand how customers evaluate the severity of a service failure. It is our recommendation that restaurant managers identify and compile a list of “typical” mistakes that customers deem most severe and focus on communicating ways to avoiding them. It is also recommended that restaurant managers develop a system for tracking and identifying incidents of severe service failure as an additional step toward ultimately avoiding severe service failures.

A critical responsibility for all front-of-house restaurant managers is to “put out multiple fires,” or effectively address multiple service failures simultaneously (or nearly simultaneously). Depending on the number of patrons and staff during a given shift, this task can be accomplished only if certain service failures are given priority over others. In addition, service failures perceived to be severe have been found to require more immediate attention (Betts et al., 2011) and additional service recovery steps (Lin, 2011). Thus, staff training should emphasize a hierarchy or continuum of service failures—identifying the most severe to least severe service failures common to the restaurant. Ongoing staff training sessions should also include a review of the hierarchy, as well as the service recovery strategy per service failure type. Finally, visual aids placed in the back of house should reinforce both the identification of a service failure (in terms of severity) and the recovery steps prescribed. For future research, a more comprehensive examination and measurement of service failures should include a determination of service failures and their severity as perceived by restaurant patrons, as well as individualized “best practices” service recovery strategies according to the severity of the service failure.

Last but not least, the results of this study demonstrate that the relative importance of the three justice dimensions depends on the type of restaurant. The findings suggest that quick-service restaurant patrons are less tolerant of an imperfect recovery (one or more justice dimensions unfulfilled) than full-service restaurant patrons. In addition, our study found that the responsiveness of the resolution is less critical for the success of recovery in the context of full-service restaurants. Based on these results, it is recommended that fast-food managers stress the importance of service recovery by providing intensive employee training focused on service recovery and developing incentives for employees who demonstrate strong recovery skills. It is also recommended that while fast-food restaurant management should emphasize the importance of responsiveness (procedural justice) to their employees, full-service restaurants should focus more on fulfilling interactive justice through kindness and empathy and providing compensation (distributive justice) commiserate with the failed service recovery attempt.

The current study has limitations that provide directions for future research. First, the study utilized a scenario-based experiment. Although this method is widely used in service recovery research, it may weaken respondents’ emotional reactions to the service failures and recoveries compared with a “real” consumption situation. Future research should focus on empirical validation of this study. Second, the manipulation of distributive injustice only involved the absence/presence of compensation. Customers may react differently to different levels of compensation. Future research should explore how the effectiveness of different levels of compensation depth. Similarly, the manipulation of interactional injustice only involved the absence/presence of an apology, in which empathy might not be fully realized. Future research can consider examining how customers’ six principles (i.e., honesty, friendliness, politeness, bias, sensitivity, and interest; Clemmer, 1993) affect the success of a service recovery attempt. Finally, our research focused on two types of restaurants: quick-service and full-service restaurants. Future research should include identifying how customers’ expectations vary across restaurants that fall between the fast food and full service. To provide additional actionable recommendations for restaurant managers, future research should focus on identifying the severity of service failures, as perceived by potential restaurant patrons. In other words, additional research is needed to develop a metric, via a ranking system or a Likert-type scale indicating importance, that can help to determine distinctions of severity among common service failures.

Footnotes

Appendix

Samples of Experiment Scenario.

(Baseline condition with high failure severity in a full-service restaurant context)
You and a group of friends go out to a full-service restaurant of your choice. The meal begins well as everyone at your table is laughing and having a great time. As you begin to take a third bite of your entrée, you see what looks to be a piece of glass on your dish. When you complain to the server, she apologizes and begins to take the dish away. Within 10 min, a fresh entrée is placed in front of you. At the end of the meal, your meal was taken off the bill.

(Procedural injustice condition with high failure severity in a full-service restaurant context)
You and a group of friends go out to a full-service restaurant of your choice. The meal begins well as everyone at your table is laughing and having a great time. As you begin to take a third bite of your entrée, you see what looks to be a piece of glass on your dish. When you complain to the server, she apologizes and begins to take the dish away. After 30 min, a fresh entrée is placed in front of you. At the end of the meal, your meal was taken off the bill.

(Interactional injustice condition with low failure severity in a quick-service restaurant context)
You and a group of friends go out to a fast-food restaurant of your choice. The meal begins well as everyone at your table is laughing and having a great time. As you begin to eat, you realize that you have been served the wrong food. When you complain to the server, she takes the dish away without an apology. Within 10 min, a fresh entrée is placed in front of you. At the end of the meal, your meal was taken off the bill.

(Distributive injustice condition with low failure severity in a quick-service restaurant context)
You and a group of friends go out to a fast-food restaurant of your choice. The meal begins well as everyone at your table is laughing and having a great time. As you begin to eat, you realize that you have been served the wrong food. When you complain to the server, she apologizes and begins to take the dish away. Within 10 min, a fresh entrée is placed in front of you. However, no discount or compensation is offered.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, or publication of this article.

ORCID iD

Suiwen (Sharon) Zou

Author Biographies

Suiwen (Sharon) Zou, PhD, is an assistant professor in the Department of Recreation, Sport and Tourism at the University of Illinois at Urbana-Champaign. As a service marketing researcher, Dr. Zou’s research focuses on service recovery, pricing strategies, and financial sustainability of tourism and leisure services.

Steven J. Migacz, PhD, is an assistant professor in the Department of Hospitality and Tourism Management at Roosevelt University. With over a decade of senior management experience in the field of hospitality, Dr. Migacz’s research focuses on food tourism and service recovery.

References

Accenture. (2016). Digital disconnect in customer engagement. https://newsroom.accenture.com/news/us-companieslosing-customers-as-consumers-demand-more-human-interactionaccenture-strategy-study-finds.htm

Amazon’s Mechanical Turk. (2014). Service summary. https://requester.mturk.com/tour

American Express. (2017). American express 2017 customer service barometer. https://business.americanexpress.com/sg/~/media/Files/GCP/sg2/business-trends-insights/Amex_GCSB_Infographics.pdf?la=en-GB

Babin

B. J.

Lee

Y. K.

Kim

E. J.

Griffin

(2005). Modeling consumer satisfaction and word-of-mouth: Restaurant patronage in Korea. Journal of Services Marketing, 19(3), 133–139.

Basso

Pizzutti

(2016). Trust recovery following a double deviation. Journal of Service Research, 19(2), 209–223.

Betts

T. K.

Wood

M. S.

Tadisina

S. K.

(2011). The impact of failure severity, prior failure, and company control on service recovery outcomes. Journal of Leadership & Organizational Studies, 18(3), 365–376.

Bitner

M. J.

Booms

B. H.

Tetreault

M. S.

(1990). The service encounter: Diagnosing favorable and unfavorable incidents. Journal of Marketing, 54(1), 71–84.

Byun

Jang

S. S.

(2015). Effective destination advertising: Matching effect between advertising language and destination type. Tourism Management, 50, 31–40.

Cai

(2018). Customers’ perceived justice, emotions, direct and indirect reactions to service recovery: Moderating effects of recovery efforts. Journal of Hospitality Marketing & Management, 27(3), 323–345.

10.

Canziani

B. F.

Almanza

McKieg

M. J.

Sullivan-Reid

(2016). Classifying restaurants to improve usability of restaurant research. International Journal of Contemporary Hospitality Management, 28(7), 1467–1483.

11.

Casado-Díaz

A. B.

Más-Ruiz

F. J.

Kasper

(2007). Explaining satisfaction in double deviation scenarios: The effects of anger and distributive justice. International Journal of Bank Marketing, 25(5), 292–314.

12.

Casado-Díaz

A. B.

Nicolau

J. L.

Mas

F. J.

(2008). The negative effects of failed service recoveries (No. 2008-07). Instituto Valenciano De Investigaciones Económicas, SA (Ivie).

13.

Choi

B. J.

(2014). The effects of perceived service recovery justice on customer affection, loyalty, and word-of-mouth. European Journal of Marketing, 48(1/2), 108–131.

14.

Chuang

Cheng

Chang

Yang

(2012). The effect of service failure types and service recovery on customer satisfaction: A mental accounting perspective. Service Industries Journal, 32(2), 257–271.

15.

Clemmer

1993. “An Investigation into the Relationship of Fairness and Customer Satisfaction with Services.” In Justice in the Workplace: Approaching Fairness in Human Resource Management, edited by Baron

, 193–207. Hillsdale, NJ: Lawrence Erlbaum.

16.

De Matos

C. A.

Henrique

J. L.

Alberto Vargas Rossi

(2007). Service recovery paradox: A meta-analysis. Journal of Service Research, 10(1), 60–77.

17.

De Matos

C. A.

Henrique

J. L.

de Rosa

(2013). Customer reactions to service failure and recovery in the banking industry: The influence of switching costs. Journal of Services Marketing, 27(7), 526–538.

18.

Dutta

Venkatesh

Parsa

H. G.

(2007). Service failure and recovery strategies in the restaurant sector. International Journal of Contemporary Hospitality Management, 19(5), 351–363.

19.

Gelbrich

Roschk

(2011). A meta-analysis of organizational complaint handling and customer responses. Journal of Service Research, 14(1), 24–43.

20.

Jang

S. S.

(2009). Perceived justice in service recovery and behavioral intentions: The role of relationship quality. International Journal of Hospitality Management, 28(3), 319–327.

21.

Haj-Salem

Chebat

J. C.

(2013). The double-edged sword: The positive and negative effects of switching costs on customer exit and revenge. Journal of Business Research, 67(6), 1106–1113.

22.

Hazée

Van Vaerenbergh

Armirotto

(2017). Co-creating service recovery after service failure: The role of brand equity. Journal of Business Research, 74, 101–109.

23.

Hoffman

K. D.

Kelley

S. W.

Rotalsky

H. M.

(1995). Tracking service failures and employee recovery efforts. Journal of Services Marketing, 9(2), 49–61.

24.

Hrazdil

Trottier

Zhang

(2013). A comparison of industry classification schemes: A large sample study. Economics Letters, 118(1), 77–80.

25.

Hur

Jang

(2016). Toward service recovery strategies: The role of consumer-organization relationship norms. Journal of Services Marketing, 30(7), 724–735.

26.

Joireman

Grégoire

Devezer

Tripp

T. M.

(2013). When do customers offer firms a “second chance” following a double deviation? The impact of inferred firm motives on customer revenge and reconciliation. Journal of Retailing, 89(3), 315–337.

27.

Jones

D. L.

McCleary

K. W.

Lepisto

L. R.

(2002). Consumer complaint behavior manifestations for table service restaurants: Identifying sociodemographic factors, personality, and behavioral factors. Journal of Hospitality & Tourism Research, 26(2), 105–123.

28.

Kim

T. T.

Jung-Eun Yoo

Lee

(2012). Post-recovery customer relationships and customer partnerships in a restaurant setting. International Journal of Contemporary Hospitality Management, 24(3), 381–401.

29.

Kim

T. T.

Kim

W. G.

Kim

H. B.

(2009). The effects of perceived justice on recovery satisfaction, trust, word-of-mouth, and revisit intention in upscale hotels. Tourism Management, 30(1), 51–62.

30.

Kim

W. G.

Lee

Y.-K.

Yoo

Y.-J.

(2006). Predictors of relationship quality and relationship outcomes in luxury restaurants. Journal of Hospitality & Tourism Research, 30(2), 143–169.

31.

Choi

(2012). The role of customer affection and trust in loyalty rebuilding after service failure and recovery. The Service Industries Journal, 32(1), 105–125.

32.

Choi

(2019). Perceived justice and CSR after service recovery. Journal of Services Marketing, 33(2), 206–219.

33.

Levesque

T. J.

McDougall

G. H.

(2000). Service problems and recovery strategies: An experiment. Canadian Journal of Administrative Sciences/Revue Canadienne Des Sciences de l’administration, 17(1), 20–37.

34.

Lin

W. B.

(2011). Construction of a service failure severity and recovery model. Expert Systems with Applications, 38(10), 12221–12230.

35.

Magnini

V. P.

Ford

J. B.

Markowski

E. P.

Honeycutt

E. D.

Jr. (2007). The service recovery paradox: Justifiable theory or smoldering myth? Journal of Services Marketing, 21(3), 213–225.

36.

Mason

Jones

Benefield

Walton

(2016). Building consumer relationships in the quick service restaurant industry. Journal of Foodservice Business Research, 19(4), 368–381.

37.

Mason

Suri

(2012). Conducting behavioral research on Amazon’s Mechanical Turk. Behavior Research Methods, 44(1), 1–23.

38.

Mattila

A. S.

(1999). An examination of factors affecting service recovery in a restaurant setting. Journal of Hospitality & Tourism Research, 23(3), 284–298.

39.

Mattila

A. S.

(2001). The effectiveness of service recovery in a multi-industry setting. Journal of Services Marketing, 15(7), 583–596.

40.

Maxham

J. G.

III. Netemeyer

R. G.

(2002). A longitudinal study of complaining customers’ evaluations of multiple service failures and recovery efforts. Journal of Marketing, 66(4), 57–71.

41.

Michel

Bowen

Johnston

(2009). Why service recovery fails: Tensions among customer, employee, and process perspectives. Journal of Service Management, 20(3), 253–273.

42.

Migacz

S. J.

(2018). Customer service recovery in hospitality: Toward an understanding of the role of the service provider utilizing Rawls’s justice theory [Doctoral dissertation]. https://oaktrust.library.tamu.edu/handle/1969.1/174002

43.

Migacz

S. J.

Zou

Petrick

J. F.

(2018). The “terminal” effects of service failure on airlines: Examining service recovery with justice theory. Journal of Travel Research, 57(1), 83–98.

44.

Namkung

Jang

(2010). Service failures in restaurants: Which stage of service failure is the most critical? Cornell Hospitality Quarterly, 51(3), 323–343.

45.

Nguyen

D. T.

McColl-Kennedy

J. R.

Dagger

T. S.

(2012). Matching service recovery solutions to customer recovery preferences. European Journal of Marketing, 46(9), 1171–1194.

46.

Nikbin

Hyun

S. S.

(2015). An empirical study of the role of failure severity in service recovery evaluation in the context of the airline industry. Review of Managerial Science, 9(4), 731–749.

47.

Nikbin

Marimuthu

Hyun

S. S.

(2016). Influence of perceived service fairness on relationship quality and switching intention: An empirical study of restaurant experiences. Current Issues in Tourism, 19(10), 1005–1026.

48.

Noone

B. M.

Kimes

S. E.

Mattila

A. S.

Wirtz

(2007). The effect of meal pace on customer satisfaction. Cornell Hotel and Restaurant Administration Quarterly, 48(3), 231–244.

49.

Oliver

R. L.

(1997). Satisfaction: A behavioral perspective on the consumer. McGraw-Hill Companies.

50.

Ong

(2012). The perceived influence of user reviews in the hospitality industry. Journal of Hospitality Marketing & Management, 21(5), 463–485.

51.

Park

Kim

O’Neill

(2014). Complaint behavior intentions and expectation of service recovery in individualistic and collectivistic cultures. International Journal of Culture, Tourism and Hospitality Research, 8(3), 255–271.

52.

Patterson

P. G.

Cowley

Prasongsukarn

(2006). Service failure recovery: The moderating impact of individual-level cultural value orientation on perceptions of justice. International Journal of Research in Marketing, 23(3), 263–277.

53.

Piaralal

S. K.

Bhatti

M. A.

Piaralal

N. K.

Juhari

A. S.

(2016). Factors affecting service recovery performance and customer service employees. International Journal of Productivity and Performance Management, 65(7), 898–924.

54.

Ponnam

Balaji

M. S.

(2014). Matching visitation-motives and restaurant attributes in casual dining restaurants. International Journal of Hospitality Management, 37, 47–57.

55.

Rawls

(1971). A theory of justice. Harvard University Press.

56.

Ring

Tkaczynski

Dolnicar

(2016). Word-of-mouth segments: Online, offline, visual or verbal? Journal of Travel Research, 55(4), 481–492.

57.

Rozin

Haidj

McCauley

Imada

(1997). Disgust: Preadaptation and the cultural evolution of a food-based emotion. Food Preferences and Taste, 65–82.

58.

Salem

Nabil

Shehata

(2017). Linking consumer characteristics to word-of-mouth-related behaviors and referral intentions in restaurants. Tourism and Traveling, 1(1), 34–50.

59.

Shin

Casidy

Mattila

A. S.

(2018). Service recovery, justice perception, and forgiveness: The “other customers” perspectives. Services Marketing Quarterly, 39(1), 1–21.

60.

Siu

N. Y. M.

Zhang

T. J. F.

Yau

C. Y. J.

(2013). The roles of justice and customer satisfaction in customer retention: A lesson from service recovery. Journal of Business Ethics, 114(4), 675–686.

61.

Smith

A. K.

Bolton

R. N.

(1998). An experimental investigation of customer reactions to service failure and recovery encounters: Paradox or peril? Journal of Service Research, 1(1), 65–81.

62.

Sparks

Fredline

(2007). Providing an explanation for service failure: Context, content, and customer responses. Journal of Hospitality & Tourism Research, 31(2), 241–260.

63.

Hsu

M. K.

Swanson

(2017). The effect of tourist relationship perception on destination loyalty at a world heritage site in China: The mediating role of overall destination satisfaction and trust. Journal of Hospitality & Tourism Research, 41(2), 180–210.

64.

Susskind

Viccari

(2011). A look at the relationship between service failures, guest satisfaction, and repeat-patronage intentions of casual dining guests. Cornell Hospitality Quarterly, 52(4), 438–444.

65.

Swanson

S. R.

Hsu

M. K.

(2011). The effect of recovery locus attributions and service failure severity on word-of-mouth and repurchase behaviors in the hospitality industry. Journal of Hospitality & Tourism Research, 35(4), 511–529.

66.

Tax

S. S.

Brown

S. W.

Chandrashekaran

(1998). Customer evaluations of service complaint experiences: Implications for relationship marketing. Journal of Marketing, 62(2), 60–76.

67.

Van Vaerenbergh

Orsingher

(2016). Service recovery: An integrative framework and research agenda. The Academy of Management Perspectives, 30(3), 328–346.

68.

Vázquez-Casielles

Suárez-Álvarez

del Río-Lanza

A. B.

(2013). The word of mouth dynamic: How positive (and negative) WOM drives purchase probability: An analysis of interpersonal and non-interpersonal factors. Journal of Advertising Research, 53(1), 43–60.

69.

Wen

Chi

C. G. Q.

(2013). Examine the cognitive and affective antecedents to service recovery satisfaction: A field study of delayed airline passengers. International Journal of Contemporary Hospitality Management, 25(3), 306–327.

70.

Weun

Beatty

S. E.

Jones

M. A.

(2004). The impact of service failure severity on service recovery evaluations and post-recovery relationships. Journal of Services Marketing, 18(2), 133–146.

71.

Yoo

J. J. E.

Shin

S. Y.

Yang

I. S.

(2006). Key attributes of internal service recovery strategies as perceived by frontline food service employees. International Journal of Hospitality Management, 25(3), 496–509.