Consumer Attributions of Product Failure to Channel Members and Self: the Impacts of Situational Cues

ABSTRACT - We investigate how situational factors such as brand visibility, product price, and problem severity affect consumer attributions of product failure to multiple parties involved in the purchase and consumption process, namely the manufacturer, the retailer and consumers themselves. We also explore the effect of gender on consumer attribution. A four-way factorial experiment using a between-subject design was conducted. The ANOVA results demonstrate the main effects of brand visibility and problem severity on consumer attribution to the manufacturer. As to consumer attribution to the retailer, we find a significant three-way interaction of brand visibility, problem severity, and gender. There is also a significant three-way interaction of brand visibility, product price, and problem severity regarding self-attribution. These results provide relevant implications for channel members in managing consumers’ reaction to product failure.

Citation:

Wanru Su and Michael J. Tippins (1998) ,"Consumer Attributions of Product Failure to Channel Members and Self: the Impacts of Situational Cues", in NA - Advances in Consumer Research Volume 25, eds. Joseph W. Alba & J. Wesley Hutchinson, Provo, UT : Association for Consumer Research, Pages: 139-145.

Advances in Consunmer Research Volume 25, 1998      Pages 139-145

CONSUMER ATTRIBUTIONS OF PRODUCT FAILURE TO CHANNEL MEMBERS AND SELF: THE IMPACTS OF SITUATIONAL CUES

Wanru Su, Yuan-Ze University

Michael J. Tippins, University of Nebraska-Lincoln

ABSTRACT -

We investigate how situational factors such as brand visibility, product price, and problem severity affect consumer attributions of product failure to multiple parties involved in the purchase and consumption process, namely the manufacturer, the retailer and consumers themselves. We also explore the effect of gender on consumer attribution. A four-way factorial experiment using a between-subject design was conducted. The ANOVA results demonstrate the main effects of brand visibility and problem severity on consumer attribution to the manufacturer. As to consumer attribution to the retailer, we find a significant three-way interaction of brand visibility, problem severity, and gender. There is also a significant three-way interaction of brand visibility, product price, and problem severity regarding self-attribution. These results provide relevant implications for channel members in managing consumers’ reaction to product failure.

INTRODUCTION

There has been extensive research on customer dissatisfaction and complaining behavior resulting from the deficiency of the product or service. Only a limited amount of attention, however, has been paid to the issue of how consumers attribute product failure to the multiple parties involved in the purchase/consumption process (Folkes and Kotsos 1986; O’Malley 1996). The key parties that could be held responsible for the product failure include the manufacturer who produces the product, the retailer who sells the product to consumers, and consumers themselves. These studies in general treat the "sellers" or "channel members" as a combined category of the retailer and the manufacturer instead of two separate parties. In addition, situational cues characterizing the specific context of product failure have not been incorporated into the existing attribution studies, although the episode-based nature of consumer reaction to product failure is recognized (Day 1980). In this research we intend to expand the scope of previous studies by investigating how situational factors such as brand visibility, price level, and problem severity affect consumer attributions toward the three major parties (namely the manufacturer, the retailer, and the customer) involved in the purchase/consumption process. This study bears important theoretical and managerial implications. A deeper understanding of how consumers place blame on multiple parties will enhance our ability to predict consumers’ complaining behavior and enrich the existing theoretical framework. Such knowledge will also allow manufacturers and retailers to address customer dissatisfaction and complaints more proactively and effectively. More importantly, manufacturers and retailers will be able to know not only how to alleviate consumer dissatisfaction, but also how to avoid shouldering unnecessary blame for other members in the channel.

LITERATURE REVIEW

Oliver (1981) argues that consumer satisfaction may best be understood as "an evaluation of the surprise inherent in a product acquisition and/or consumption experience" (p.27). He proposes a comprehensive model of the satisfaction process in retail settings, which consists of three evaluation stages: purchase experience in the store, product consumption, and redress activities. The model indicates that the satisfaction with the store affects the satisfaction with product consumption, and both in turn affect the satisfaction with redress activities. This model, however, is unable to offer any guidance regarding how situational cues in a specific product purchase/consumption experience influence customer reaction to product failure, as well as their subsequent impact on customer satisfaction and complaining behavior.

Among alternative research approaches to customer satisfaction and complaining behavior, the attributional approach in particular has great potential for offering valuable insights regarding how situational factors influence consumer reaction to product failure (see Folkes [1988] for a review). This approach emphasizes the culpable source(s) of a disconfirmation (i.e., whose fault it is), not just the existence of a disconfirmation. Most attribution studies on product failure in the marketing literature focus on self-attribution (e.g., Folkes 1984) by examining the three causal dimensions of attribution: locus of control, stability, an controllability (Weiner 1986). Folkes and Kotsos (1986) take a step further to compare the customer’s and the retailer’s attribution patterns. They asked the respondents to decide to what extent "the buyer" and "the sellers" (which was a combined category of the retailer and the manufacturer in their study) should be blamed respectively for the incident of product failure described in a scenario. They find that people in general tend to blame the sellers for the problem rather than the buyer, with some self-serving bias on the part of the retailers when the scenario deals with their own products. O’Malley (1996), on the other hand, develops a conceptual model predicting the adoption of central and peripheral processing routes in consumer attribution of product failure, as well as emotion and behavioral consequences toward the channel members. One limitation of both studies is the focus on the traditional dichotomy of internal-external attribution. A further differentiation of external sources of product failure between the manufacturer and the retailer can provide useful insights into the coordination of channel relationship, especially the management of channel conflict due to product failure.

Another limitation of the existing attribution literature on product failure is the lack of due attention to the effects of situational factors on consumer attribution. Day (1980) argues that customer complaints and redress-seeking actions are motivated by specific aspects of experiences with particular products or services; among those "episode-specific" variables are cost/benefit evaluation, attribution of blame, probability of successful redress, and the type of product or service involved in the situation of dissatisfaction. We further propose that attribution of blame can also be affected by these situational factors. Since in most cases the causes of product failure are not clear-cut, "blame placement" among multiple potential culprits are inevitably subject to the influence of situational cues such as product characteristics and the purchase/consumption experience. Attribution of blame, along with other episode-specific variables, will then in turn affect the customer satisfaction and complaining behavior.

In this study we examine how the specific situational cues affect consumer attributions of product failure to the manufacturer, the retailer, and self. We choose to begin our investigation with three situational cues: brand visibility, product price, and problem severity. Both brand visibility and product price are considered two of the most immediate aspects of product purchase/consumption experience while problem severity reflects the reduction of consumption utility due to product failure. Given the fact that products are often physically handled by both the manufacturer and the retailer, and to some extent by the customer before the final purchase, the causes of product deficiencies may not be readily evident. Such ambiguity of blame allocation provides us with an ideal opportunity to examine how these situational cues interact with one another in determining the attributions of product failure among possible culprits.

Effect of Brand Visibility

Brand visibility could influence consumer attribution in different ways. On one hand, higher visibility could make the manufacturer a more salient target in the mind of consumers. According to the salience or primacy rule of consumer accepting the first adequate explanation (Kelley and Michela 1980), the highly visible manufacturer thus may receive more blame than the little known one. On the other hand, high brand visibility may shield the manufacturer from receiving blame by projecting a better quality image. The resulting high-quality image makes the manufacturer a less plausible cause of product failure, leaving the retailer as a more likely culprit instead. Thus two sets of competing hypotheses regarding the effect of brand visibility are developed accordingly. The first set of hypotheses assume te adoption of the salience rule while the second set of hypotheses are based on the dominance of the quality-inference rule.

Hypothesis 1a: The manufacturer receives more blame for a high-visibility brand than a low-visibility brand when the product fails.

Hypothesis 1b: The retailer receives less blame for a high-visibility brand than a low-visibility brand when the product fails.

Hypothesis 1c: The customer receives less blame for a high-visibility brand than a low-visibility brand when the product fails.

Hypothesis 2a: The manufacturer receives less blame for a high-visibility brand than a low-visibility brand when the product fails.

Hypothesis 2b: The retailer receives more blame for a high-visibility brand than a low-visibility brand when the product fails.

Hypothesis 2c: The customer receives more blame for a high-visibility brand than a low-visibility brand when the product fails.

Effect of Product Price

Given the generally accepted positive correlation between price and quality, a high price may reduce the amount of blame placed on the manufacturer by signaling high product quality. As a result, the retailer is more likely to receive blame for not inspecting the product thoroughly before giving it to the customer while charging the latter a high price. A low price, on the contrary, may indicate low quality; consumers may tend to blame themselves in the face of product failure for consciously making the price-quality tradeoff at the time of purchasing. Thus the following hypotheses are developed with respect to the effect of product price.

Hypothesis 3a: The manufacturer receives less blame for a high-priced product than a low-priced product when the product fails.

Hypothesis 3b: The retailer receives more blame a high-priced product than a low-priced product when the product fails.

Hypothesis 3c: The customer receives less blame for a high-priced product than a low-priced product when the product fails.

It should be noted that the effect of product price could be moderated by involvement rather than price-quality inference. Goodman et al. (1995) demonstrate the moderating effect of involvement. Given the dissatisfaciton with central transaction factors, they find that the customers with greater involvement show higher overall dissatisfaction with their suppliers. In the case of product purchase/consumption, greater involvement with a more expensive item may influence not only consumer satisfaction but also consumer attribution of product failure. Unlike the attribution based on price-quality inference, higher involvement may generate more causal elaboration, through which more possible culprits are identified and less blame is placed on a single party. As a result, the size of the changes predicted in Hypotheses 3a to 3c may be diluted to some extent.

Effect of Problem Severiy

In a survey of consumer reactions to common purchases, Best and Andreasen (1977) find household status and problem type influence the perceptions of problems and the choice of action or inaction. Of particular interest here is their finding that the very nature of the deficiency in a product or service affects whether people perceive the deficiency as a problem. That is, manifest or clear-cut shortcomings of the purchase are easy to identify and acknowledge while ambiguous or complicated deficiencies are relatively difficult to perceive clearly and state with assurance. In our study, we approach the impact of product deficiency in a product failure situation from the angle of problem severity. When encountering a major (or serious) problem with the product, consumers experience a significant reduction in the utility of consuming the product due to weaker performance or less appeal. For instance, the product may turn out to be dysfunctional, or become less attractive or pleasant to use. In the case of a minor problem, however, the reduction of consumption utility is negligible and of small personal consequence. In many cases, a major problem tends to be relatively easier to trace back to an unambiguous culprit (and is likely to be an external source) while the cause for a minor problem is less clear-cut. Thus the following hypotheses are developed accordingly:

Hypothesis 4a: The manufacturer receives more blame for a major problem than a minor problem when the product fails.

Hypothesis 4b: The retailer receives more blame for a major problem than a minor problem when the product fails.

Hypothesis 4c: The customer receives less blame for a major problem than a minor problem when the product fails.

Effect of Gender

The differences between males and females regarding their attributions for success and failure in academic or professional situations have been well documented (see the review in Kelly and Michela 1980). For example, Ryckman and Peckham (1987) show that as compared to boys, girls are more likely to ascribe their success to unstable factors such as luck or effort, and their failure to stable factors such as task difficulty or ability. Similarly, Stipek and Gralinski (1991) show that girls expect less success, rate their ability lower, are less likely to attribute success to ability and failure to luck, and are more likely to attribute failure to low ability. Russo (1991) examines the gender differences of adults in self and social attributions for career success. They find that females perceive the combined factor of ability and hard work as more important to their own success than that of male colleagues. In addition, self-attributions to luck and self-attribution to ability and hard work are positively correlated for females, but are negatively correlated for males. It seems that luck and effort play a stronger role in female adults’ attribution of their own success as compared to male adults.

It is not clear how these findings about gender difference in the existing attribution literature can be generalized from a context of academic or professional achievement to that of product purchase and consumption. There is little research on gender difference in the attribution of product failure in the marketing literature. No formal hypothesis is formulated here regarding the effect of gender on customer attribution. It is our attempt to explore the gender effect on consumer attribution and use the preliminary findings as a basis for future research.

METHODOLOGY

Design and Stimuli Material

We conduct a four-way factorial experiment using a between subject design with two treatment levels for each factor in the experiment. The three major independent variables include brand visibility, product price, and problem severity, while gender is used as blocking a factor. We conducted pretests with college students regarding their experiences with multiple products and retail stores. We decide to use the product category of "athletic shoes" in the scenarios of the experiment because the pretest scores indicate relatively high levels of experience and interest regarding the purchase of athletic shoes among college students. In light of the common criticism that studies using college students as a convenience sample lack proper generalizability, we make an effort to minimize this problem by choosing a product category for which college students, along with teenagers, are considered the dominant group of buyers. We also decided to use the same national specialty store chain across all the scenarios in order to control the variance of the retail sources that respondents may refer to in their decision task. According to the pretest scores, the selected specialty store projects an image of higher quality and greater service orientation as compared to discount stores and mass merchandisers. Thus we may not be able to generalize our findings to lower-end retail stores.

Among the three independent variables, brand visibility is operationalized as a popular national brand versus a fictitious brand. As to product price, we select $24.99 and $89.99 as the low and high price conditions after checking the price range of athletic shoes at the particular specialty store in the local area. Problem severity is operationalized as a major problem ("a deep scratch in the leather) versus a minor problem ("a frayed shoelace"). Both the specific price levels and product problems were tested with college students in advance to ensure sufficient perceived differences between conditions of the experiment. Eight scenarios are constructed accordingly. Each scenario describes an incident of product failure with detailed information about the price, the brand, and- the deficiency of the product. The respondents are told that they bought the product at a particular store and find out the problem after they brought it back home.

There are three dependent measures of consumer attribution for product failure in this experiment. Respondents are asked to indicate the extent to which each of the three parties ("Yourself," "Manufacturer," and "Retailer") should be held responsible for the problem described in the scenario. These dependent variables are measured in a seven-point scale with 1 being "no responsibility" and 7 being "fully responsible"; thus a higher number indicates a greater amount of blame placed on a party. As a manipulation check, we also measure the perceptions of respondents toward the brand, the price, and the problem described in the scenario along a seven-point scale, with a higher number indicating a more visible brand, a more expansive product, or a more serious problem. Additional measures such as respondents' experience, interest, and knowledge levels regarding the purchase of athletic shoes are included as well.

Subjects and Procedures

In total two hundred and thirty seven undergraduate students participated in the experiment for extra course credit. The experiment was conducted in seven separate sessions in a normal classroom setting. In each session, the surveys were randomly distributed to the students after a brief introduction about the task. Each survey contains only one scenario. Students were instructed to read the scenario and decide to what extent each of the three parties involved in the purchase/consumption process should be held responsible for the problem described in the scenario. They were then asked to indicate their perceptions about the visibility of the brand, the price level, and the degree of seriousness of the problem described in the scenario. Students concluded the task by describing their experience of and interest in buying athletic shoes and how knowledgeable they were about purchasing athletic shoes.

SUMMARY OF RESULTS AND DISCUSSION

The results of the manipulation check indicate that the manipulation of various conditions in the experiment is successful. The average perceptions toward the visibility of the national brand (mean=6.73), the high-price level (mean=4.67), and the severity of the major problem (mean=4.82) are all significantly higher than the neutral level of the seven-point scale (p<0.00). Similarly, the average perceptions toward the visibility of the fictitious brand (mean=1.53), the low price level (mean=1.71), and the severity of the minor problem (mean=2.72) are all significantly lower than the neutral level of the seven-point scale (p<0.00). Overall, respondents perceive the national brand to be more visible than the fictitious brand (difference of means=5.20, F(1, 235)=1623.88, p <0001), the high-price level to be more expensive than the low-price level (difference of means=2.96, F(1, 235)=420.03,p <0001) and the major problem to be more serious than the minor problem (means=2.10, F(1, 235)=144.78, p <.0001).

Respondents in general show higher levels of experience (mean=5.37), interest (mean=4.36), and knowledge (mean=4.92) than the neutral level of the seven-point scale (p<0.002). Male respondents, however, show higher self-ratings of experience (difference of means=0.63, F(1, 235)=11.69,p <001) and knowledge (difference of means=0.76, F(1, 235)=17.46,p <0001) than female respondents.

We perform separate four-way full factorial ANOVA procedures regarding consumer attributions of product failure to the manufacturer, the retailer, and self. Table 1 summarizes the respective ANOVA results. Table 2 includes the means and standard deviations of the three dependent variables for the overall sample, the male sub-sample, and the female sub-sample.

Consumer Attribution of Product Failure to the Manufacturer

With respect to the blame placed on the manufacturer, we find the significant main effects of brand visibility (F(1, 221)=5-92, p <02) and problem severity (F(1, 221)=4.87,p <03), but fail to find any strong impact of product price and gender. The results show that the manufacturer of the fictitious brand receives more blame (mean=5.17) than the manufacturer of the national brand (mean=4.65). Our result supports Hypothesis 2a and suggests the dominance of the quality-inference route over the salience route in the attribution process. We also find that the manufacturer receives more blame for the minor problem (mean=5.15) than for the major problem (mean=4.67). This finding, however, is opposite to the prediction of Hypothesis 4a.

TABLE 1

SUMMARY OF ANOVA RESULTS

Consumer Attribution of Product Failure to the Retailer

As to the blame placed on the retailer, the ANOVA results indicate a significant three-way interaction of brand visibility, problem severity, and gender (F(1, 221)=6.61,p <02), but there is no significant effect of product price. Further analyses for each brand show a significant simple interaction effect of problem severity and gender for the fictitious brand (F(1, 117)=11-01, P=.001). Separate analyses for each gender group under the fictitious-brand condition indicate a significant simple main effect of problem severity for females (F(1, 53)=7.66,p<0-01), and a marginally significant simple main effect of problem severity for males (F(1, 64)=3.53, p--0.065).

The above results show that when the product is a national brand, product price, problem severity, and gender all have no impact on consumer attribution of product failure to the retailer. In the case of the fictitious brand, however, males and females reveal different attribution patterns. Specifically, females place more blame on the retailer for a major problem (mean=5.70) than for a minor problem (mean=4.61). Thus Hypothesis 4b is partially supported. Males, on the contrary, tend to place more blame on the retailer for a minor problem (mean=5.50) than for a major problem (mean=4.84), though the difference is only marginally significant in the statistical sense.

Consumer Attribution of Product Failure to Self

For the blame consumers placed on themselves, we find a significant three-way interaction of brand visibility, price, and problem severity (F(1, 221)=4.89, p <.03), but did not find any significant effect of gender. With further analyses at the brand level, we find a significant simple main effect of price for the fictitious brand (F(1, 117)=5.97, p <.02), and a significant simple interaction effect of price and problem severity for the national brand (F(1, 112)=5.36, p <.03). With respect to the national brand, additional analyses show that problem severity has a significant simple main effect at the low price level (F(1, 56)=4.56,p<0.04), but not at the high price level (F(1, 56)=1.35, p=0.251).

These results show that in the case of a fictitious brand, people tend to blame themselves more for a low-price item (mean=3.52) than for a high-price item (mean=2.79) regardless of the problem they encounter. This finding provides partial support to the effect of product price predicted in Hypothesis 3c regarding customer attribution of product failure to self. In the case of the national brand, however, consumer attribution seems to depend on the price level. For a low-price national brand, people tend to blame themselves more for a minor problem (mean=3.80) than for a major problem (mean=2.86). For a high-price national brand, however, problem severity has no impact on self-attribution. Thus Hypothesis 4c regarding the effect of problem severity on self-attribution can hold only under the condition of a low-priced national brand.

TABLE 2

MEANS OF THE BLAME PLACED ON THE MANUFACTURER, THE RETAILER AND SELF

GENERAL DISCUSSION

Our results show that when encountering product failure, consumers are more likely to blame the manufacturer of a little-known brand than that of a highly visible brand. This finding suggests the customers' tendency to infer high product quality from high brand visibility, which in turn helps the manufacturer shift the responsibility of product failure to other parties. In addition, we find that people tend to blame the manufacturer more for a minor problem than for a major problem. One possible explanation for this finding is that people may believe a major problem such as a deep scratch in the leather is less likely to occur at the manufacturer's production site; instead it is more likely to occur through consumer trials in the retail store. On the contrary, a minor problem such as a frayed shoelace can occur either in the production process or in the store since both the manufacturer and the retailer have the opportunity to handle the shoelaces. In addition, given the nature of the product involved (i.e., athletic shoes), customers may believe that a major problem is less likely to pass the manufacturer's quality control than a minor problem. More importantly, customers may believe that the retailer should handle the shoes with care and do the final inspection of the shoes thoroughly.

As to the attribution of product failure to the retailer, we find that males and females reveal different patterns of attribution in the case of a fictitious brand. Males tend to place more blame on the retailer for a minor problem while females tend to place more blame on the retailer for a major problem. One possible explanation is that drawing from their knowledge and experience of purchasing athletic shoes, males believe a minor problem such as a frayed shoelace is more likely to occur in the retail store than a major problem such a deep scratch in the leather. Females, on the other hand, report less experience and knowledge regarding the purchase of athletic shoes than males do; therefore, females may be more inclined to take the retailer as the salient target to blame for a major problem than for a minor problem.

Finally, we find that product price and brand visibility affect the extent to which consumers are willing to take (or share) the blame for product failure. In the case of a little known brand, consumers tend to blame themselves more for a low-price item than for a high-price item regardless of the severity of the problem. This seems to suggest that consumers are blaming themselves for knowingly and willingly making the price-quality tradeoff (i.e., taking chances on the low-price item while being aware of its low or unstable quality) at the time of purchasing. In the case of a well-known brand purchased at a low price, consumers tend to blame themselves more for a minor problem than for a major problem. Since consumers are more likely to overlook a minor problem such as a frayed shoelace than a major problem such as a deep scratch in the leather, they may tend to blame themselves more for such negligence.

CONCLUSION

Our study provides empirical evidence to support the importance of situational factors in their influence over consumer attribution of product failure. We also expand the existing literature by examining how consumers attribute the product failure to each of the major parties involved in the purchase/consumption process, including the retailer, the manufacturer, and consumers themselves. We focus on three specific situational factors-brand visibility, product price, and problem severity--and demonstrate -the differential effects of these factors on consumer attributions to all three parties. In addition, we find that male and female consumers have different attribution patterns toward the retailer.

In addition to the limited generalizability due to the use of a student sample, there are three major limitations of our study. First, we recognize that the effect of problem severity (a "minor" versus "major" problem) might be confounded with that of problem type "frayed shoelace" versus "deep scratch") in our experiment. A more stringent manipulation of problem severity would require the control for the problem type. In the case of athletic shoes, for instance, we can use "a small scratch" versus "a deep cut" to manipulate problem severity while keeping both conditions in the same problem category of "surface damage." Second, both of the product problems included in the experiment are related to the appearance of the product instead of its physical performance. We choose the appearance-related problems for this study mainly because all three parties (the manufacturer, the retailer, and the customer) have the opportunity to physically handle the product and thus can all be plausible causes of the deficiency in product appearance. It is important, however, to see if the results can hold if the problems are performance-related, especially for the manufacturer. Finally, we did not measure the underlying processes of customer attribution in this experiment. This severely limits our ability to offer direct evidence to explain the specific processes through which the three situational cues affect consumer attribution. While observing the effects of these situational cues, we can only offer inferences regarding why such effects take place.

We believe further replications and extensions of this research are needed to enrich the existing theoretical framework of consumer attribution. We need to better understand how situational factors influence consumer attribution of product failure, dissatisfaction level, and complaining behavior. It is necessary to extend the current study to other types of product problems, product categories and retail stores. We also plan to measure and analyze the reasoning process behind consumer attributions to multiple parties along the three dimensions (locus, controllability, and stability) identified in Weiner's (1986) attribution theory. Both verbal protocols and specific measures could be used to capture the process. It is also important to investigate the impacts of other situational factors such as store image, awareness of available redress options (i.e., refund or free exchange policies), the amount of handling needed in the retail store, and the technological complexity of the product. With such enhanced knowledge about consumer attribution, manufacturers and retailers will be able to effectively alleviate or even avoid the blame that consumers place upon them when encountering product failure.

REFERENCES

Best, Arthur, and Alan R. Andreasen (1977), "Consumer Response to Unsatisfactory Purchases: A Survey of Perceiving Defects, Voicing Complaints, and Obtaining redress," Law & Society, 11 (Spring), 701-742.

Day, Ralph L. (1980), "Research Perspectives on Consumer Complaining Behavior," in Lamb and Dunne (eds.), Theoretical Developments in Marketing, Chicago, IL: American Marketing Association, 211-215.

Folkes, Valerie S. (1984), "Consumer Reactions to Product Failure: An Attributional Approach," Journal of Consumer Research, 10 (March), 398-409.

Folkes, Valerie S. (1988), "Recent Attribution Research in Consumer Behavior: A Review and New Directions," Journal of Consumer Research, 14 (March), 548-565.

Folkes, Valerie S. And Barbara Kotsos (1986), "Buyers' and Sellers' Explanations for Product Failure: Who Done it? Journal of Marketing, 50 (April), 74-80.

Goodman, Paul S., Mark Fichman, F. Javier Lerch, Pamela R. Snyder (1995), "Customer-Firm Relationships, Involvement, and Customer Satisfaction," Academy of Management Journal, 38, 1310-1324.

Kelley, H. H., and J. L. Michela (1980), "Attribution Theory and Research," Annual Review of Psychology, 31, 475-501.

Oliver, Richard L. (1981), "Measurement and Evaluation of Satisfaction Processes in Retail Settings," Journal of Retailing, 57 (Fall), 25-48.

O'Malley, John R. Jr. (1996), "Consumer Attribution of Product Failures to Channel Members," Advances of Consumer Research, 23, 342-345.

Ryckman, D., and R. Peckham (1987), "Gender Differences in Attributions for Success or Failure," Journal of Early Adolescence, 7, 47-63.

Russo, Nancy Felipe, Rita Mae Kelly, and Melinda Deacon (1991), "Gender and Success-related Attributions: Beyond Individualistic Conceptions of Achievement," Sex Roles: A Journal of Research, 25 (September), 331-350.

Stipek, D. J., and J. H. Gralinski (1991), "Gender Differences in Children's Achievement-related Beliefs and Emotional Responses to Success and Failure in Mathematics," Journal of Educational Psychology, 83, 361-371.

Weiner, Bernard (1986), An Attributional Theory of Motivation and Emotion, New York: Springer-Verlag.

----------------------------------------