CoChat
Sign Up
Cited by 176Customer Service Quality and LoyaltyService and Product InnovationAI in Service Interactions

Recovery Voice and Satisfaction After Service Failure: An Experimental Investigation of Mediating and Moderating Factors

Open access
Karande, Kiran, Magnini, Vincent P., Tam, Leona·2007·Journal of Service Research
DOI 10.1177/1094670507309607
Open this paper in CoChat — free accountOpen PDF

Abstract

Past research studies on service failures and recovery have conceptualized “voice” in terms of customers having an opportunity to air complaints after failures occur. In contrast, the authors introduce the concept of recovery voice , which entails a service provider asking a customer (after a failure has occurred) what the firm can do to rectify the problem. In a scenario-based experiment carried out in an airline setting and in a hotel setting with 216 and 208 participants, respectively, it was found that customers perceived greater procedural justice when offered recovery voice, which resulted in higher overall postfailure satisfaction. It was shown that perceived procedural justice mediated the effect of recovery voice on overall satisfaction. Furthermore, recovery voice had a greater impact on perceived procedural justice for established customers with long transaction histories than for new ones with short transaction histories. Managerial and research implications based on these findings are also presented.

AI snapshot

Written from the paper’s full text.

TL;DR

The paper introduces recovery voice—asking customers how a firm should fix a service failure—and tests it in airline and hotel scenarios. It finds recovery voice raises perceived procedural justice, which mediates higher postfailure satisfaction, with past transaction history moderating the first link but gender not moderating either link.

Key findings

  • Across 216 airline and 208 hotel participants, recovery voice increased perceived procedural justice (combined M = 6.14 vs. 4.83; F(1, 416) = 60.04, p < .01), and higher procedural justice increased overall postfailure satisfaction (combined high vs. low M = 6.78 vs. 4.10; F(1, 416) = 264.20, p < .01).
  • Procedural justice mediated recovery voice’s effect on satisfaction: Sobel tests were significant in airline, hotel, and combined samples (4.14, 4.50, 5.88, all p < .01); Baron and Kenny’s procedure indicated perfect mediation for airline and partial mediation for hotel.
  • Past transaction history moderated the recovery voice → procedural justice link: the effect was stronger for established/long-history customers than for new/short-history customers (combined interaction F(1, 416) = 7.61, p < .01; simple effects F = 55.49 vs. 15.91, p < .001).
  • Gender did not moderate recovery voice → procedural justice or procedural justice → satisfaction (Hypotheses 4 and 5 not supported); Hypothesis 7, that procedural justice → satisfaction would be stronger for new customers, was not supported, though means were in the predicted direction.

Methods

  • A 2 × 2 × 2 between-subjects scenario experiment manipulated recovery voice (present/absent) and past transaction history (long/short), with gender measured; separate airline and hotel service-failure vignettes were used.
  • Data came from an online consumer panel: 424 responses total, 216 for airline scenarios and 208 for hotel scenarios; participants were aged 25–82 (average 53).
  • Overall postfailure satisfaction was measured with three 9-point semantic-differential items (α = .96 combined), and procedural justice with six 9-point Likert-type items (α = .95 combined); analysis used ANOVA, simple-effects tests, Baron and Kenny mediation steps, and Sobel tests.

Limitations

  • The authors state this is the first study of recovery voice and caution that findings should be interpreted carefully; recovery voice may have negative effects if customer input is not followed or is unrealistic, and asking for input may make customers question provider competence.
  • The authors note that online vignettes cannot fully capture feelings surrounding real service failures, past experience was held constant, only airline and hotel settings were studied, hotel recovery was immediate while airline recovery occurred the next day, and service-category differences were significant.
  • Evident from the text: the sample was an online panel rather than actual customers experiencing failures; hotel mediation was only partial, leaving a direct recovery voice effect unexplained; and several hypotheses, especially gender moderation and Hypothesis 7, were unsupported.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1177/1094670507309607

Cites

55 of 55 references verified0 citation errors · 0 not checkable

Statistics

18 tests recomputed · 0 with the reported and recomputed p on opposite sides of .05

Claims the paper makes

0 of 53 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    Perceived procedural justice mediated the effect of recovery voice on overall postfailure satisfaction.

    “It was shown that perceived procedural justice mediated the effect of recovery voice on overall satisfaction.”
    Quote in the paperSubstantial claim
  • Findingown data

    Recovery voice had a greater impact on perceived procedural justice for established customers with long transaction histories than for new customers with short transaction histories.

    “recovery voice had a greater impact on perceived procedural justice for established customers with long transaction histories than for new ones with short transaction histories.”
    Quote in the paperSubstantial claim
  • Findingown data

    Recovery voice resulted in higher overall postfailure satisfaction.

    “which resulted in higher overall postfailure satisfaction”
    Quote in the paperSubstantial claim
  • Findingown data

    In the scenario-based experiments, customers offered recovery voice perceived greater procedural justice than customers not offered recovery voice.

    “it was found that customers perceived greater procedural justice when offered recovery voice”
    Quote in the paperSubstantial claim
  • Findingown data

    Medium to large effect sizes of procedural justice on overall postfailure satisfaction were found (partial η2 = .15, .19, and .19 for the airline, hotel, and combined samples, respectively).

    “Medium to large effect sizes of procedural justice were found (partial η 2 = .15,.19,and .19 for the airline, hotel, and combined samples, respectively)”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The items used to measure procedural justice, overall satisfaction, recovery voice, and past transaction history are valid measures of those constructs.
  • May combine two claims. 22
  • Classified as probably not a substantial claim. 32

  • May combine two claims.

    Findingown data

    In the hotel scenario, perceived procedural justice means for established customers were 6.37 without recovery voice and 7.72 with recovery voice, and for new customers were 4.95 without recovery voice and 7.14 with recovery voice.

    “For the hotel scenario, perceived procedural justice means for established customers were 6.37 without recovery voice and 7.72 with recovery voice, F(1, 200) = 31.69,p < .001,and for new customers were 4.95 without recovery voice and 7.14 with recovery voice, F(1, 200) = 6.11, p < .05.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The items used to measure procedural justice, overall satisfaction, recovery voice, and past transaction history are valid measures of those constructs.

  • May combine two claims.

    Findingown data

    In the airline scenario, perceived procedural justice means for established customers were 3.99 without recovery voice and 5.66 with recovery voice, and for new customers were 3.35 without recovery voice and 4.12 with recovery voice.

    “The airline scenario showed similar results with perceived procedural justice: Means for established customers were 3.99 without recovery voice and 5.66 with recovery voice, F(1, 208) = 21.56,p < .001,and for new customers were 3.35 without recovery voice and 4.12 with recovery voice, F(1, 208) = 5.15, p < .05).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The items used to measure procedural justice, overall satisfaction, recovery voice, and past transaction history are valid measures of those constructs.

  • May combine two claims.

    Findingown data

    Recovery voice was consistently and significantly correlated with procedural justice and overall satisfaction in both the airline and hotel scenarios.

    “It can be seen that recovery voice was consistently and significantly correlated with procedural justice and overall satisfaction in both the airline and hotel scenarios.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The items used to measure procedural justice, overall satisfaction, recovery voice, and past transaction history are valid measures of those constructs.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Although Hypothesis 7 was not supported, the means were in the hypothesized direction.

    “Thus, although Hypothesis 7 was not supported, the means were in the hypothesized direction.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Interaction effects between recovery voice and past transaction history on procedural justice were significant for the hotel sample and the combined sample, and approached significance for the airline sample.

    “Interaction effects between recovery voice and past transaction history on procedural justice were significant for the hotel sample, F(1, 200) = 4.86, p < .05,and the combined sample, F(1, 416) = 7.61, p < .01,and approached significance for the airline sample, F(1, 208) = 3.84, p = .07).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The items used to measure procedural justice, overall satisfaction, recovery voice, and past transaction history are valid measures of those constructs.

  • May combine two claims.

    Findingown data

    Procedural justice had a significant positive effect on overall postfailure satisfaction for the airline scenario, the hotel scenario, and the combined sample.

    “Consistent with Hypothesis 2, customers had higher overall postfailure satisfaction when they had higher procedural justice perceptions (M = 5.93, 7.04, and 6.78 for the airline, hotel, and combined samples, respectively) than lower procedural justice perceptions (M = 3.82, 5.10, and 4.10 for the airline, hotel, and combined samples, respectively) across all samples.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The items used to measure procedural justice, overall satisfaction, recovery voice, and past transaction history are valid measures of those constructs.

  • May combine two claims.

    Findingown data

    The combined analysis indicates that the service category has a significant impact on both perceived procedural justice and overall postfailure satisfaction.

    “the combined analysis indicates that the service category has a significant impact on both perceived procedural justice and overall postfailure satisfaction.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Granting a customer voice enables the customer to serve as a cocreator in a service encounter.

    “Likewise, granting a customer voice enables the customer to serve as a cocreator in a service encounter and enhances the customer's satisfaction should a failure occur later (Mattila and Cranage 2005).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Mattila and Cranage 2005 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Encouraging customers to voice their complaints typically enhances customer satisfaction ratings.

    “Within the context of service recovery, earlier research has found that encouraging customers to voice their complaints typically enhances customer satisfaction ratings (Sparks and McColl-Kennedy 2001).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Sparks and McColl-Kennedy 2001 source

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    Overall satisfaction is an aggregation of all previous transaction-specific evaluations and is updated after each service encounter.

    “From a different perspective, overall satisfaction is an aggregation of all previous transaction-specific evaluations and is updated after each service encounter (Boulding et al. 1993), like a moving average (Parasuraman, Zeithaml, and Berry 1994).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Boulding et al. 1993 source
    • cites Parasuraman, Zeithaml, and Berry 1994 source

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    Customer reactions to service failure depend on what is done to fix problems (outcomes), as well as how they are resolved (processes).

    “Many service recovery studies have pointed out that customer reactions to service failure depend on what is done to fix problems (outcomes), as well as how they are resolved (processes) (e.g., Mattila and Patterson 2004;Sparks and McColl-Kennedy 2001).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Mattila and Patterson 2004 source
    • cites Sparks and McColl-Kennedy 2001 source

  • May combine two claims.

    Premisecitation

    Women tend to be more process oriented and to show greater interest in how they are treated by service providers.

    “Women, on the other hand, tend to be more process oriented and to show greater interest in how they are treated by service providers, in terms of whether the service providers are concerned about what happened to them and whether they allow them to voice their concerns (McColl-Kennedy, Daus, and Sparks 2003).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites McColl-Kennedy, Daus, and Sparks 2003 source

  • Classified as probably not a substantial claim.

    Study design and conductown data

    Respondents were randomly assigned to one of the eight treatments (four hotel and four airline).

    “Those responding to the invitation were randomly assigned to one of the eight treatments (four hotel and four airline).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationreasoning

    The difference in the strength of mediation between the hotel and airline scenarios may be attributed to the difference in recovery speeds presented in the scenarios.

    “The difference in the strength of mediation between the hotel and airline scenarios may be attributed to the difference in recovery speeds presented in the scenarios.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    An outcome better than expected results in positive disconfirmation, leading to satisfaction.

    “An outcome of the evaluation that is poorer than prior expectations results in a negative disconfirmation of expectations, which leads to dissatisfaction, whereas an outcome that is better than expected results in a positive disconfirmation of expectations leading to satisfaction.Confirmation occurs when performance matches expectations and does not lead to a change in satisfaction.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Churchill and Surprenant 1982 source

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    Confirmation occurs when performance matches expectations and does not lead to a change in satisfaction.

    “An outcome of the evaluation that is poorer than prior expectations results in a negative disconfirmation of expectations, which leads to dissatisfaction, whereas an outcome that is better than expected results in a positive disconfirmation of expectations leading to satisfaction.Confirmation occurs when performance matches expectations and does not lead to a change in satisfaction.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Churchill and Surprenant 1982 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Participants in the airline and hotel scenarios had similar demographic characteristics.

    “Participants in the airline and hotel scenarios had similar demographic characteristics.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    The mean realism scores for the four scenarios for each service category were high and at about the same level.

    “Table 1c shows that the mean realism scores for the four scenarios for each service category were high and at about the same level.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The written scenarios realistically represent actual service failure situations, such that participants' responses reflect how real customers would respond.

  • May combine two claims.

    Premisecitation

    For customers with satisfactory past experiences or customers who are loyal, expectations for recovery tend to be high.

    “Thus, for customers with satisfactory past experiences or customers who are loyal, expectations for recovery tend to be high, which makes the process of complaint handling highly important (Kelley and Davis 1994;Tax, Brown, and Chandrashekaran 1998).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Kelley and Davis 1994 source
    • cites Tax, Brown, and Chandrashekaran 1998 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Perceptions of procedural justice positively affect satisfaction with a variety of variables such as pay, performance appraisals, layoff policies, and selection procedures.

    “studies in organizational research have found that perceptions of procedural justice positively affect satisfaction with a variety of variables such as pay, performance appraisals, layoff policies, and selection procedures (Greenberg 1990).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Greenberg 1990

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    Persons given opportunities to express their views believe that they have greater control over the outcomes, which leads to greater procedural fairness judgments.

    “That is, the voice effect is explained by the presumption that persons given opportunities to express their views believe that they have greater control over the outcomes, which leads to greater procedural fairness judgments (Lind, Kanfer, and Early 1990).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Lind, Kanfer, and Early 1990 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Poor complaint handling will damage the relationship most when prior experience is positive and expectations for recovery are high.

    “Consistent with these arguments, the service quality literature has also found that poor complaint handling will damage the relationship most when prior experience is positive and expectations for recovery are high (Tax, Brown, and Chandrashekaran 1998).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Tax, Brown, and Chandrashekaran 1998 source

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    An outcome poorer than prior expectations results in negative disconfirmation, which leads to dissatisfaction.

    “An outcome of the evaluation that is poorer than prior expectations results in a negative disconfirmation of expectations, which leads to dissatisfaction, whereas an outcome that is better than expected results in a positive disconfirmation of expectations leading to satisfaction.Confirmation occurs when performance matches expectations and does not lead to a change in satisfaction.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Churchill and Surprenant 1982 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Perceptions of justice affect overall satisfaction with the service encounter.

    “In addition, in the services literature, it has been found that perceptions of justice affect the overall satisfaction with the service encounter (Mattila and Cranage 2005;Smith, Bolton, and Wagner 1999;Sparks and McColl-Kennedy 2001;Tax, Brown, and Chandrashekaran 1998).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Mattila and Cranage 2005 source
    • cites Smith, Bolton, and Wagner 1999 source
    • cites Sparks and McColl-Kennedy 2001 source

  • Classified as probably not a substantial claim.

    Premisecitation

    A customer's satisfaction is the consequence of an evaluation process in which the customer judges his or her expectations of how a service should be performed against the actual service experience.

    “The moderating effect of past transaction history in the procedural justice-overall satisfaction relationship is supported by the expectancy disconfirmation paradigm, which posits that a customer's satisfaction is the consequence of an evaluation process in which the customer judges his or her expectations of how a service should be performed against the actual service experience (Oliver 1989).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Oliver 1989

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationreasoning

    For new customers, granting recovery voice increases their perceptions of procedural justice, which exceed their expectations, leading to their satisfaction.

    “On the other hand, for new customers, granting recovery voice increases their perceptions of procedural justice, which exceed their expectations, leading to their satisfaction.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    New customers are not familiar with the provider because they lack transaction histories and therefore have less defined expectations.

    “In contrast, new customers are not familiar with the provider because they lack transaction histories and therefore have less defined expectations (Solomon et al. 1985).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Solomon et al. 1985 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Judgments of the fairness of a decisionmaking procedure are enhanced when the individual involved is offered an opportunity to express his or her views and opinions before the decision is made.

    “This argument can be supported by the consensus in the psychology literature that judgments of the fairness of a decisionmaking procedure are enhanced when the individual involved in the situation is offered an opportunity to express his or her views and opinions before the decision is made (Lind, Kanfer, and Early 1990).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Lind, Kanfer, and Early 1990 source

  • May combine two claims.

    Interpretationreasoning

    For established customers, granting recovery voice results in high perceived procedural justice that meets their expectations and therefore maintains their existing satisfaction.

    “For established customers, granting recovery voice results in high perceived procedural justice that meets their expectations and therefore maintains their existing satisfaction.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Granting a customer voice enhances the customer's satisfaction should a failure occur later.

    “Likewise, granting a customer voice enables the customer to serve as a cocreator in a service encounter and enhances the customer's satisfaction should a failure occur later (Mattila and Cranage 2005).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Mattila and Cranage 2005 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    Hotel stays tend to be more pleasant than airline travel.

    “This might be because hotel stays tend to be more pleasant than airline travel (American Consumer Satisfaction Index 2007a, 2007b).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites American Consumer Satisfaction Index 2007a, 2007b

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisenone given

    In past studies, voice was operationalized as the opportunity to voice a complaint, whereas in our study, it is an opportunity to offer input into how the service failure should be resolved.

    “In past studies, voice was operationalized as the opportunity to voice a complaint, whereas in our study, it is an opportunity to offer input into how the service failure should be resolved.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Regression analysis using perceived procedural justice as a continuous independent variable found identical results.

    “We also carried out regression analysis using perceived procedural justice as a continuous independent variable and found identical results.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationreasoning

    Perceptions of incompetence may result from asking for customer input in a recovery situation.

    “Moreover, perceptions of incompetence may result from asking for customer input in a recovery situation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given

    Offering a consumer a voice in the recovery situation increases the expectation that the input is listened to.

    “For example, offering a consumer a voice in the recovery situation increases the expectation that the input is listened to.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Past research has operationalized voice as the willingness of a customer to air his or her complaint to a firm.

    “Past research has operationalized voice as the willingness of a customer to air his or her complaint to a firm (e.g., Didow and Barksdale 1982;Goodwin and Ross 1992;Sparks and McColl-Kennedy 2001).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Didow and Barksdale 1982 source
    • cites Goodwin and Ross 1992 source
    • cites Sparks and McColl-Kennedy 2001 source

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    If the consumer's demands are deemed unrealistic and are not fully acted on, the end result could be negative.

    “Thus, if the consumer's demands are deemed unrealistic and are not fully acted on, the end result could be negative.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    The experiment entailed responding to online vignettes, so feelings and emotions surrounding actual service failures were not fully experienced.

    “Because the experiment entailed responding to online vignettes, feelings and emotions surrounding actual service failures were not fully experienced.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Others, such as Smith, Bolton, and Wagner (1999), have modeled the effect of recovery attributes such as response speed on procedural justice and procedural justice on satisfaction separately but have not specifically tested for its mediating effect.

    “Others, such as Smith, Bolton, and Wagner (1999), have modeled the effect of recovery attributes such as response speed on procedural justice and procedural justice on satisfaction separately but have not specifically tested for its mediating effect.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Smith, Bolton, and Wagner (1999) source

  • Classified as probably not a substantial claim.

    Premisecitation

    When an interaction effect is present, main effects should not be interpreted directly but examined using simple effects.

    “According to Keppel and Wickens (2006), when an interaction effect is present, the main effects should not be interpreted directly but examined using simple effects, because a significant interaction effect means that the effect of one independent variable depends on another variable.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Keppel and Wickens (2006)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Gender was expected to make a difference on the basis of similar results found in the psychology literature in pay and performance appraisals, layoff policies, and selection procedures.

    “We had expected gender to make a difference on the basis of similar results found in the psychology literature in pay and performance appraisals, layoff policies, and selection procedures (Greenberg 1990).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Greenberg 1990

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    Asking consumers for input in constructing their redress initiatives can result in negative outcomes.

    “First, it is prudent to note that asking consumers for input in constructing their redress initiatives can result in negative outcomes.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    The scenario approach enables costly and difficult manipulations to be more easily operationalized.

    “This scenario approach enables costly and difficult manipulations to be more easily operationalized”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Past experience was held constant across all scenarios.

    “In our study, past experience was held constant across all scenarios.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisereasoning

    The mediation test is a stronger test of how the impact of recovery voice on overall postfailure satisfaction takes place.

    “The mediation test is a stronger test of how the impact of recovery voice on overall postfailure satisfaction takes place.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Customers vary in their importance to service firms.

    “Specifically, as suggested by that literature, customers vary in their importance to service firms, and their individual value should therefore be assessed (e.g., Bell et al. 2002;Hogan, Lemon, and Libai 2003;Venkatesan and Kumar 2004).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Bell et al. 2002 source
    • cites Hogan, Lemon, and Libai 2003 source
    • cites Venkatesan and Kumar 2004 source
46 of 46

  • The paper relies on this without support.

    The written scenarios realistically represent actual service failure situations, such that participants' responses reflect how real customers would respond.

    “The extent to which the respondents found the scenarios to be realistic was measured by asking the question”

  • The paper relies on this without support.

    The items used to measure procedural justice, overall satisfaction, recovery voice, and past transaction history are valid measures of those constructs.

    “The respondents indicated their perceptions of recovery voice on two items:”

  • The paper relies on this without support.

    The airline and hotel service failure scenarios are comparable enough to combine into a single sample for analysis.

    “The same ANOVA models were also estimated with the combined sample using the same set of independent and dependent variables.”

  • The paper relies on this without support.

    The online consumer panel respondents are representative of consumers who experience service failures.

    “Data were collected using an online consumer panel of a large reputed marketing research company.”
4 of 4

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Justice perceptions completely mediate the relationship between organizational responses and post-complaint satisfaction.

    “It is argued and empirically shown that justice perceptions (distributive, procedural, and interactional justice) completely mediate the relationship between organizational responses and post-complaint satisfaction (Karande, Magnini, and Tam 2007; Maxham III and Netemeyer 2003; Smith, Bolton, and Wagner 1999).”

    From A Meta-Analysis of Organizational Complaint Handling and Customer Responses (2011) · cites this paper as Karande, Magnini, and Tam 2007

    The Mediation Analysis reports that procedural justice fully mediated the effect of recovery voice on postfailure satisfaction in the airline scenario, but only partially mediated it in the hotel scenario. The study did not test distributive or interactional justice as mediators, or organizational responses beyond recovery voice; the broader claim of complete mediation is therefore not supported. (LLM, conf=0.94)

    Experimental — accuracy not yet measured

Statistics that don’t add up

Each reported p-value recomputed from its test statistic and degrees of freedom.

  • Inconsistent p-value: F(1, 208) = 3.84, p = .07

    Minor

    The paper reports p = .07 here; recomputing from the test statistic and its degrees of freedom gives p = .051. The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p = .07 · recomputed p = .051

1 of 1

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Doesn’t add upF(1, 208) = 3.84reported p = .07 · recomputed p = .051

    “F(1, 208) = 3.84, p = .07”

  • ConsistentF(1, 208) = 22.24reported p < .01 · recomputed p < .001

    “F(1, 208) = 22.24, p < .01”

  • ConsistentF(1, 200) = 31reported p < .01 · recomputed p < .001

    “F(1, 200) = 31.00,p < .01”

  • ConsistentF(1, 416) = 60.04reported p < .01 · recomputed p < .001

    “F(1, 416) = 60.04,p < .01”

  • ConsistentF(1, 208) = 53.27reported p < .01 · recomputed p < .001

    “F(1, 208) = 53.27,p < .01”

  • ConsistentF(1, 200) = 54.16reported p < .01 · recomputed p < .001

    “F(1, 200) = 54.16,p < .01”

  • ConsistentF(1, 416) = 264.2reported p < .01 · recomputed p < .001

    “F(1, 416) = 264.20,p < .01”

  • ConsistentF(1, 200) = 4.86reported p < .05 · recomputed p = .029

    “F(1, 200) = 4.86, p < .05”

  • ConsistentF(1, 416) = 7.61reported p < .01 · recomputed p = .006

    “F(1, 416) = 7.61, p < .01”

  • ConsistentF(1, 416) = 55.49reported p < .001 · recomputed p < .001

    “F(1, 416) = 55.49,p < .001”

  • ConsistentF(1, 416) = 15.91reported p < .001 · recomputed p < .001

    “F(1, 416) = 15.91,p < .001”

  • ConsistentF(1, 208) = 21.56reported p < .001 · recomputed p < .001

    “F(1, 208) = 21.56,p < .001”

  • ConsistentF(1, 208) = 5.15reported p < .05 · recomputed p = .024

    “F(1, 208) = 5.15, p < .05”

  • ConsistentF(1, 200) = 31.69reported p < .001 · recomputed p < .001

    “F(1, 200) = 31.69,p < .001”

  • ConsistentF(1, 200) = 6.11reported p < .05 · recomputed p = .014

    “F(1, 200) = 6.11, p < .05”

  • ConsistentF(1, 416) = 3.93reported p < .05 · recomputed p = .048

    “F(1, 416) = 3.93, p < .05”

  • ConsistentF(1, 416) = 179.14reported p < .001 · recomputed p < .001

    “F(1, 416) = 179.14,p < .001”

  • ConsistentF(1, 416) = 79.57reported p < .001 · recomputed p < .001

    “F(1, 416) = 79.57,p < .001”

18 of 18

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Transaction-history interaction F statistics differ between text and table

    Major

    For the recovery-voice-by-transaction-history interaction, the results text reports F = 4.86 for hotels and 3.84 for airlines; Table 4a reports 6.13 and 3.15, respectively.

  • Reported high-justice satisfaction means exceed both gender-specific means

    Major

    Table 3 gives high-procedural-justice satisfaction means of 5.04 and 3.70 for airline men and women, so their aggregate cannot be 5.93. The hotel aggregate of 7.04 likewise exceeds both listed gender means, 7.00 and 5.47.

  • Final summary names the wrong relationship for Hypothesis 6

    Minor

    Hypothesis 6 concerns recovery voice's effect on procedural justice, not its effect on overall satisfaction as the summary states.

  • Recovery-voice group sizes disagree between procedure and Table 1a

    Minor

    The procedure assigns 109 airline and 97 hotel respondents to recovery voice; Table 1a instead labels those groups 107 and 111. The corresponding no-voice counts also disagree.

  • Hotel transaction-history group sizes are reversed in Table 1b

    Minor

    The procedure reports 111 hotel participants with short histories and 97 with long histories. Table 1b assigns those counts to long and short histories, respectively.

  • Airline realism cells account for only 207 participants

    Minor

    The four airline cell counts in Table 1c total 207, rather than the reported airline sample of 216.

    Reported 216 · recomputed 207

  • Discussion reverses the setting comparison for transaction-history correlations

    Minor

    The discussion says the correlations involving transaction history are higher for hotels. Table 2 instead lists higher airline correlations with procedural justice (.27 versus .02) and satisfaction (.33 versus .04).

7 of 7

Cites

Flagged first, then by citations

Research works printed but not linked: 10

  1. American Consumer Satisfaction Index, 2007, Scores by Industry: Airlines
  2. 2007, Scores by Industry: Hotels
  3. Keppel, Geoffrey, 2006, Design and Analysis: A Research's Handbook
  4. 1976, Fairness in Social Responsibilities
  5. Tom R. Tyler, 1988, The Sociology Psychology of Procedural Justice
  6. Nunnally, Jum C., 1994, Psychometric Theory
  7. 1989, Journal of Customer Satisfaction, Dissatisfaction, and Complaining Behavior, 2, 146
  8. Rosener, Judy B., 1990, Harvard Business Review, 68, 119
  9. Sparks, Beverley A., 2001, Journal of Business, 54, 209
  10. Tax, Stephen S., 1998, Journal of Marketing,, 62, 6

Cited by

Sorted by citations

Related

Authors
Karande, K., Magnini, V. P., & Tam, L.
Affiliations
  • Old Dominion University (US)
  • Longwood University (US)
ORCID
Venue
Journal of Service Research
Publisher
SAGE Publications
Published
2007-11
Volume
10
Issue
2
Pages
187-203
ISSN
  • 1094-6705 (ISSN-L)
  • 1552-7379
DOI
10.1177/1094670507309607
OpenAlex
W2148799739
Cited by
176
References
52
Citations by year
2026: 5 · 2025: 4 · 2024: 5 · 2023: 7 · 2022: 5 · 2021: 17 · 2020: 3 · 2019: 13 · 2018: 10 · 2017: 11 · 2016: 12 · 2015: 16 · 2014: 14 · 2013: 11 · 2012: 8
FWCI
13.63
Type
journal article
Language
en
Open access
Licence
journals.sagepub.com/page/policies/text-and-data-mining-license (tdm)
Locations
  • Journal of Service Research (publishedVersion)
  • Digital Commons - Longwood (Longwood University) (submittedVersion)
Topics
  • Social Sciences › Business, Management and Accounting › Organizational Behavior and Human Resource Management › Customer Service Quality and Loyalty
  • Social Sciences › Business, Management and Accounting › Marketing › Service and Product Innovation
  • Physical Sciences › Computer Science › Artificial Intelligence › AI in Service Interactions
Keywords
service failure, procedural justice, customer complaints, service recovery, airline services, hotel services, transaction history, customer relationships

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-11.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-11.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review