An Extension to the Process of Customer Service Quality Evaluation Through Psychology and Empirical Study

ABSTRACT - There is undoubtedly a psychological basis to the process of customer service quality evaluation (CSQE). By looking at the fundamental psychology framework as a whole from the point of view of the pure psychology literature, we can identify additional suggestions to the process of CSQE. This paper reports Ph.D. research at the University of Edinburgh by the author, which suggests that the customer’s service quality evaluation, for both a service experience and a service provider, is derived by that customer using one or other of at least three comparison heuristics in order to benchmark services against each other. These suggestions build on the major previous theories of CSQE.

Citation:

Philip E. Lewis (2001) ,"An Extension to the Process of Customer Service Quality Evaluation Through Psychology and Empirical Study", in AP - Asia Pacific Advances in Consumer Research Volume 4, eds. Paula M. Tidwell and Thomas E. Muller, Provo, UT : Association for Consumer Research, Pages: 281-287.

Asia Pacific Advances in Consumer Research Volume 4, 2001      Pages 281-287

AN EXTENSION TO THE PROCESS OF CUSTOMER SERVICE QUALITY EVALUATION THROUGH PSYCHOLOGY AND EMPIRICAL STUDY

Philip E. Lewis, University of Vaasa, Finland

ABSTRACT -

There is undoubtedly a psychological basis to the process of customer service quality evaluation (CSQE). By looking at the fundamental psychology framework as a whole from the point of view of the pure psychology literature, we can identify additional suggestions to the process of CSQE. This paper reports Ph.D. research at the University of Edinburgh by the author, which suggests that the customer’s service quality evaluation, for both a service experience and a service provider, is derived by that customer using one or other of at least three comparison heuristics in order to benchmark services against each other. These suggestions build on the major previous theories of CSQE.

Whilst there is no general agreement on the nature of Perceived Service Quality, and whilst explanations are regularly changing and increasing, key researches and opinions on the topic (Parasuraman, Zeithaml and Berry 1985; Boulding, Kalra, Staelin and Zeithaml 1993; Liljander and Strandvik 1993; Liljander 1995) can arguably be combined into the following guiding base definition: Perceived Service Quality is the result of a process. It is a customer’s comparison, summarized in an attitude, of the believed goodness or badness f all dimensions of a particular service experience or service provider, in relation to expectations.

When Gr÷nroos (1982) introduced the concept of 'Perceived Service Quality’, quality was deemed good if perceptions of service at least matched the expected service standard. Gr÷nroos (1982) initially only related this evaluation to an experience on a particular occasion. Similar predictive expectations have been suggested also within the satisfaction literature relating to goods (Bolfing and Woodruff 1988; Churchill and Surprenant 1982), and services (Prakash and Lounsbury 1984; Swan 1988). Gr÷nroos’ principle was the basis of early research by Parasuraman, Zeithaml and Berry (1985) which in turn was the basis of SERVQUAL (Parasuraman, Zeithaml and Bery 1988; Parasuraman, Berry and Zeithaml 1991), a method of measurement designed to enable service providers to evaluate their customers’ perceptions of service quality. SERVQUAL, however related to the evaluation of a service provider rather than a service experience.

SERVQUAL is arguably incomplete since the theory of the nature of expectations has since been expanded to include two dynamic rather than just one post-purchase type of expectation. Boulding, Kalra, Staelin and Zeithaml (1993) argued and tested the view that individuals’ perceptions of the service quality of a firm just after a service contact are a blend of their prior expectations of what will and what should transpire during the contact, and the actual delivered service during the service encounter. This much is also echoed by Liljander and Strandvik (1993) and Liljander (1995). The Boulding et al. (1993) theory suggests there are two types of expectation. The first is a predictive expectation, 'will’ expectations and the second is a 'should’ expectation which acts as a standard for comparison of the perceived service.

Furthermore, the nature of comparison standards are arguably far more diverse than allowed for within SERVQUAL or the theory behind it. The above theories and suggestions by other researchers such as Liljander (1995), Cadotte, Woodruff and Jenkins (1983,1987), Bolfing and Woodruff (1988), Prakash and Lounsbury (1984), and Swan (1988) suggest many possible comparison standards including predicted service, ideal service, excellent service, desired service, deserved service, needs and values, cultural norms, promises, adequate service, best brand norm, brand norm, product type norm, favorite brand model, comparative expectations, equity and fairness. However, SERVQUAL only really incorporates a rough aggregated mixture of a selection of these. Liljander (1995) argues that there is not enough detail and preciseness in this use of expectations.

Therefore, if service providers are to be able to measure their customers Perceived Service Quality precisely and informatively they must incorporate a precise understanding of the nature of comparison expectations being separate from prediction expectations. The other limitations of SERVQUAL such as reliability, discriminant validity and variance restriction (Brown, Churchill and Peter 1993) should, ideally also be overcome. It is often not enough to just know if customers receive adequate service or exceed expectations because most service providers which are concerned with differentiating through service are neither providers of outstanding service, nor providers of inadequate service. For these companies this situation is unlikely to change. However, they want to be providers of service of a significantly higher level than provided by all or most of their competitors.

This article reports research which tries to find a more coherent explanation for the use of comparison expectations within the process of Perceived Service Quality. It does so by freshly looking at the problem from the point of view of findings within the pure psychology literature rather than the point of view of the psychology related business researcher. This does not mean that the business literature is ignored, on the contrary, but it means that the limitations of the conventions of Perceived Service Quality research are relaxed somewhat in the hope of finding nw insights.

In this rest of this article, perceived service quality will be referred to as 'Customer Service Quality Evaluation’ (CSQE) as first presented at the 'International and Multidisciplinary Symposium on Service Quality 1996BQuis 5’ (Lewis 1996). This wording change is not a criticism of the original Perceived Service Quality concept but a recognition that the word 'perception’ as used in psychology (Gregory 1986) is used slightly differently to the way Gr÷nroos uses it. In the field of psychology perceptions do not contain attitudes but rather contribute towards them. When asking customers to state their CSQE, CSQE is taken to be an evaluation of how good a service experience or service provider, such as a particular bank, is in the opinion of that customer.

At the end of 1996 the author’s Ph.D. thesis reported an extension to the evaluation of the CSQE concept and empirical tests for its measurement. This thesis has never been publicly released and so this is the first publicly available post completion report on the research. This expanded understanding is empirically tested to be appropriate to the banking industry in the United Kingdom and in particular large retail banks in Britain. However, it is believed appropriate to all service providers because the psychology literature used to develop the theory is not appropriate to just one industry, and because the findings of this thesis have suggested that CSQE usually takes place in the context of all service providers. This report summarizes the Ph.D.’s findings and considers ways and plans to extend this research in the near future.

Put simply, the research which this report refers to tests the suggestion that a customer’s service quality evaluation, for both a service experience and a service provider, is derived by that customer using one of at least 3 CSQE heuristics. These CSQE heuristics are achieved by the customer comparing her or his generic attitude for a service experience, or service provider, with her or his generic comparison attitudes. These comparison attitudes are comprised of attitudes for outstanding, normal, and appalling service, (top, average and worst service). The generic attitude for the service experience or service provider is also compared with four other intermediate levels of service, together with the customer believed incidence of occurrence of service experiences or service providers at each of those levels.

This use of expectations does not deny the existence of prediction expectations. On the contrary, prediction expectations are proposed both by the business and psychology literature. There is also no assumption that a customer necessarily evaluates the quality of a service experience or service provider after each service encounter. These suggestions therefore do not contradict the major previous theories of CSQE, as much as they build on them. In this way it is believed that understanding has been extended in this area of research.

THE USE OF PSYCHOLOGYBAN EXTENDED THEORY OF CSQE

We know, through simple observation, that customers often speak of how a service or service provider is better or worse than expected, better or worse than usual, could be better or worse, the best or worst, unbelievably good or bad, and relatively good or bad. Comments such as these suggest there are many reference points in memory which are used by customers, not just two reference points as suggested by the 'Zone of Tolerance’ of Berry and Parasuraman (1991). Whilst this is not of course scientifically tested evidence, we all know that every one of us have made and heard comments such as those mentioned above, on numerous occasions. Such evidence therefore, provides us with a common sense hunch, and sense of direction, concerning the nature of what is taking place in the minds of customers when evaluating the quality of service experiences or service providers. This hunch is suppoted by the logical conclusions, backed by some imagination, which can be drawn from the fundamental psychology framework.

The fundamental psychology framework suggests that following attention, perception, storage in memory and manipulation of service experiences and all other relevant interacted information, expectations in the form of schemata can be formed as a means of comparison for specific service experiences. This use of expectations does not deny the existence of prediction expectations. On the contrary, prediction expectations are proposed both by the business and psychology literature.

To be of any use however, schemata should cover the whole range of experience in order to allow a service experience to be compared relatively with all other experiences. This requires at least a schema for the upper limit (best service) and a schema for the lower limit (worst service). The distribution of service experiences between the upper and lower limit would be required however, in order to give an accurate position of a service experience within any given comparative distribution. A schema for typical service, together with a knowledge of how many experiences are similar to each schemata, is therefore beneficial.

A customer can gain knowledge of how many experiences are similar to each of the schemata, by conducting an audit of how easy it is to recall experiences of a similar nature of positiveness to the schemata. The easier it is to recall experiences, the more experiences that there are believed to be. This type of psychological process is known as the Availability Heuristic (Tversky and Kahnemann 1973; Gabrielcik and Fazio 1984). A very similar method has been shown to be used for the placing of events in chronological time (Brown, Rips and Shevell 1985). The evaluation of the number of experiences which are similar to each schemata, could also take place using the knowledge of such facts held in memory.

These comparison expectations could be made more accurate through the use of four additional points of reference: 'similar to but worse than best schema’, 'similar to but better than norm schema’, 'similar to but worse than norm schema’, 'similar to but better than worst schema’. This could be possible since typicality and familiarity are features of the perception system (Gregory 1986; Goldstein 1989) and memory systems (Kohonen 1984; Kintsch 1980). A knowledge of the number of experiences which are similar to each of these intermediate service levels would also assist accurate evaluation of service.

Since the customer knows approximately how many past direct and indirect experiences correspond to each level of service, the customer can now know approximately how many comparative experiences are bettered by the experience in question, and therefore comparatively how good the experienced service was in relation to all other service experiences. In this way a service quality evaluation can be obtained.

The logistical problem which still exists following the formation of comparison expectation schemata, is the question of how to turn the customer’s schemata, as well as memory for the service experience, into compatible units for comparison. Schemata would appear to consist of many elements. Schemata are generalizations of procedures, social events and social behaviour, and are thus cognitively and evaluatively complex (Rumelhart 1975,1980; Rumelhart and Norman 1985). The memory for a service experience is also complex. For comparison of schemata and memory for a service experience to be possible, both must be made evaluatively simple so that a single value for each of the schemata, and a single value for the memory for the experience, can be compared with each other. The only cognitively complex and evaluatively simple stores of information known to psychology, that would be relevant to this situation, are attitudes. The schemata and service experience memory therefore need to be turned into attitudes if they are to be compared.

Once the attitude towards the service experience and those for expectations have been formed, and any resultant cognitive dissonance reduce if possible, a comparison between the two can be made. This comparison, together with the knowledge of the number of service experiences similar to each of the expectations schemata, will result in the CSQE.

Another question which remains however, is the question of exactly how the comparison between the service experience and comparison attitudes takes place. A customer, in her or his mind, could place the service experience attitude as equal to the middle of a specific schemata category, believing that the service is typical, and therefore somewhere in the middle of all service experiences provided in that category. We will call this the Half CSQE Heuristic. Alternatively, a customer could place the service experience attitude as equal to the top of a schemata category, believing that the service is as good as any service experiences in that category. We will call this the Full CSQE Heuristic. As a third alternative, a customer could make a simpler, though probably less accurate CSQE, by comparing a service attitude only with the comparison schemata and intermediate service levels, rather than also with the incidence of occurrence at each of those schemata and intermediate service levels. We will call this the Basic CSQE Heuristic. Perhaps two or all three of these CSQE heuristics are used, but at different times, or by different customers. Perhaps none of these heuristics, or slightly different ones are used. This suggests that there are two scales of comparison, a full incidence scale and an ordinal scale, that can be derived from the same information, namely three schemata and four intermediate service levels.

It seems reasonable to argue that the memory system provides individuals with the ability to derive the information necessary to form an attitude towards a particular service provider. Similarly it should also be possible for an individual to derive the information from memory necessary to form schemata and attitudes about similar service providers’ service. If this is the case, then all of the elements of the process of CSQE mentioned above and illustrated in Figure 1, should also apply to CSQE of a particular service provider.

FIGURE 1

HYPOTHESES

3 key testable hypotheses were derived from the theory. Firstly, expectations, as a means of comparison for the evaluation of specific service experiences comprise 2 components: attitudes for appalling, normal, and outstanding service experiences, together with 4 intermediate levels of service, as provided by similar service providers; an estimate of how often each level of such service experiences are provided. Expectations as a means of comparison for the evaluation of generic service providers such as Barclays Bank, comprise the same two components, except that in this case the two components relate to service providers rather than to service experiences.

CSQE for a specific service experience is determined by the position of these comparison expectations relative to a customer’s attitude towards the service experience. CSQE for a service provider is determined by the position of comparison expectations relative to a customer’s attitude towards the service provider. Comparison expectations used for the evaluation of a specific service experience are termed 'specific expectations’ and those used for the evaluation of a given service provider such as a particular bank are termed 'generic expectations’. Service improvement is therefore stepwise, in that unless the incidence of occurrence changes, CSQE can only change if service is equated with a different point of reference.

The second hypothesis is that customers use one of at least 3 related heuristics (shortcuts or rules of thumb), as explained above, when comparing the service experience or service provider with 'comparison expectations’. A customer may either believe hat all service providers which relate to a chosen point of reference are bettered by the evaluated service provider (this is known as the 'Full Expectations Heuristic’), or that only half are bettered (this is known as the 'Half Expectations Heuristic’). Alternatively a customer may not consider the incidence of service providers when evaluating CSQE. In this case, CSQE will be decided purely according to the point of reference with which the customer equates the service provider’s service (this is known as the 'Basic Expectations Heuristic’).

The third hypothesis is that comparison expectations’ of customers include knowledge of a whole range of types of service providers, not just the industry that is appropriate to the service experience or service provider being evaluated. For example, when evaluating the service of a bank, a customer will compare that service with the service believed to be provided by services such as supermarkets, estate agents, solicitors and banks. These broad 'comparison expectations’ were referred to in this research as 'wide expectations’, the opposite, whereby comparison expectations refer to companies in the same industry, were referred to as 'narrow expectations’.

FIGURE 2

CORRELATION WITH CUSTOMER STATED CSQE

THE EMPIRICAL STUDY TO TEST THE THEORY

Four questionnaire versions were distributed for the empirical study in order to see if wide expectations are used rather than narrow expectations and in order to see if the proposed process of CSQE is appropriate to both specific and generic evaluations. The 4 versions of the questionnaire were: Specific-Narrow, Specific-Wide, Generic-Narrow and Generic-Wide. All four of these questionnaires were identical except for the necessary changes made to each one in order to make it relate to the different kind of evaluation. One questionnaire version concerned specific experience CSQE. It tested whether the proposed process of CSQE applies to specific service experiences of retail banks. This questionnaire version was split into two sub-versions. These two sub-version questionnaires tested if expectations are as specific as, or more than, just 'banks’ for specific CSQE. These two sub-version questionnaires were named 'Specific Narrow’ and 'Specific Wide’ respectively. Another questionnaire version concerned generic CSQE. This concerned customers evaluations of retail banks. This questionnaire version was also split into two sub-versions. These two sub-versions tested if expectations are as specific as, or more than, just 'banks’ for generic CSQE. These two sub-version questionnaires were named 'Generic Narrow’ and 'Generic Wide’.

Analysis of returned questionnaires was quantitative. Values for each of the three CSQE heuristics, for each subject were achieved by calculating CSQE in the three ways suggested by the three CSQE heuristics. Firstly in order to achieve this, the position of the service relative to best, worst and normal schemata, and four intermediate service levels, was obtained through a series of questions in questionnaires. Secondly, the proportions of experiences (or service providers in the case of generic service quality) out of ten or 100, in the minds of subjects, which were similar or equal to each of the schemata and intermediate levels, were then obtained through further questions in the questionnaires. Thirdly, from this information, the proportion of experiences less good than the experienced service (or the proportion of service providers less good than the service provider being evaluated, in the case of generic service quality) was calculated.

Full CSQE heuristic values included all of the incidence of experiences or service providers similar to the comparison schema or intermediate levels with which a subject equated service, together with the incidences for all lower schemata and intermediate levels. Half CSQE heuristic values included the same as the full CSQE heuristic, except that half o the incidence for the schemata or intermediate level with which the subject equated the service was subtracted from the prediction. Basic CSQE heuristic values were calculated by allocating an equal space interval value to each successively higher comparison schemata and intermediate level, starting at zero for appalling service, with five for normal service, ten for outstanding service, and so on for the four intermediate comparison points.

To see if any or all of the CSQE heuristics are an accurate predictor of 'Stated CSQE’, CSQE stated by the customer was correlated with each of the three CSQE heuristic values.

For each of the four questionnaire types, 1000 questionnaires were distributed to a random sample of subjects around Britain. Of the 4000 questionnaires, 36 per cent or 1440, were returned to the researchers. The proportions of questionnaires returned were very similar for each of the four questionnaire types and subjects used in the final analysis were demographically, and in other ways, extremely variable.

The findings of the empirical study suggested that customers do indeed seem to usually use one or other of the three heuristics and that wide expectations appear more likely, both during CSQE of specific experiences and during CSQE of generic service providers. The correlations for each of the heuristic predictions against actual stated CSQE values are described in Figure 2. Figure 2 suggests that overall the 3 heuristics each correlate well with CSQE but that generally heuristics using wide expectations correlate better than those using narrow expectations, though the differences are sometimes small. Furthermore, the basic heuristic correlates best of all the heuristics except when using narrow expectations to evaluate a specific service experience. It should be noted that because each heuristic is not the only possible heuristic it is surely unlikely that a much higher correlation could be found.

This evidence might suggest that perhaps consumers only use the basic heuristic and therefore no incidence of occurrence. However, the evidence becomes clearer when it is noticed that in most cases at least one of the 3 heuristic predictions is no more than 5% away from actual stated CSQE (with the specific narrow questionnaire at least one heuristic was accurate to within 5% in 52.7% of cases, with specific wide it was 58.3%, with generic narrow it was 61.8% and with generic wide it 67.5%) but that it is rare (less than 10% of cases) for more than one heuristic prediction to be accurate at the same time. This suggests that no one heuristic is sufficient for predicting CSQE.

A counter argument to the above evidence might be that if 3 predictions are spaced out at lets say 10% intervals, then one of the predictions is going to always be within 5% of CSQE so long as the range of variance of CSQE for any given ordinal level of service does not exceed 30%. However, the heuristic based predictions were accurate for a far larger degree of variance. This finding is illustrated by Figure 3. In other words, the graphs also showed that heuristics were predictive of extreme as well as moderate actual stated CSQE, even if the CSQE related to the same descriptive level of service. It is not known however, to what extent specific customers use the same or a different one of the three heuristics for different CSQEs, to what extent and why.

Sometimes none of the three heuristics appeared to be used. Once again, the reason or reasons for this are not known but perhaps it is because other heuristics may be used by some customers, or by all customers at certain times for certain reasons, or because specific customers use narrow expectations or even wider or narrower expectations, for all or some CSQEs. It would be irrational to presume that all consumers always use only one of just 3 and therefore it is not surprising that the results suggest some missing heuristics.

FIGURE 3

CONTRIBUTIONS OF THE RESEARCH TO THE UNDERSTANDING OF CSQE

A number of aspects of the theory of CSQE suggested by this research are contained in other theories of CSQE. Comparison schemata are incorporated into the process for the evaluation of generic service provider CSQE, as with the Parasuraman, Berry and Zeithaml (1990) approach; comparison schemata are also incorporated into the process for the evaluation of specific service experience CSQE, as with the Gr÷nroos (1982) approach and the Boulding, Kalra, Staelin and Zeithaml (1993) approach. Furthermore, as with both of those approaches, the schemata incorporated into the process are somewhat flexible and appropriate/realistic schemata, not just wishful thinking (i.e. they incorporate a 'zone of tolerance’). Also, as with the Boulding et al. approach and the Liljander (1995) approach, a CSQE evaluation feeds into memory/image relating to that service provider. Furthermore, this theory of CSQE does not contradict the existence of prediction expectations as suggested by Gr÷nroos (1982), Boulding et al. (1993), and Liljander (1995). Nor does it contradict any of the dimensions of service, determinants of perceptions and expectations, or sources of service interaction commonly explained by the most of the key authors in this field of research. In fact the fundamental psychology framework which backs up the theory proposed by this research (itself a contribution to understanding of the process of CSQE because it combines current satisfaction, service quality and psychology literature), suggests the integration of all the above characteristics of the process of CSQE into the extended theory of CSQE, which is empirically supported by this research.

There are, however, some significant differences between the process of CSQE suggested by this research, and the process suggested by the other theories. Firstly, this theory suggests that customers evaluate the quality of specific service experiences and generic service providers using the same kind of process. Other theories either do not consider CSQE for both specific service experiences and generic service providers, or suggest that the two evaluations use different kinds of processes. All the following contributions apply to CSQE of specific service experiences and generic service providers.

Secondly, the process suggested by this research incorporates a larger number of integrated comparison schemata than other theories, 3 rather than just 1 or even none. Furthermore, it incorporates 4 integrated intermediate points of comparison/comparison standards, unlike other theories which do not suggest any. This is not to say that a variety of comparison standards have not been considered before, they have (e.g. Liljander and Strandvik 1993; Liljander 1995).

Thirdly, the theory suggested here incorporates a knowledge, by customers, of the number of service experiences or service providers similar to each point of comparison, unlike other theories.

Fourthly, in the theory suggested here, it is an overall service provider attitude or service experience attitude that is then compared with overall comparison attitudes, for service providers or service experiences, in order to derive the CSQE. In most other theories individual perception and expectation dimensions, which are weighted according to the customer’s believed dimension importance, are compared with each other, and the comparisons for all the dimensions are combined to derive an overall CSQE.

Fifthly, the theory suggested here incorporates the use of heuristics by customers. Furthermore, the incorporation of additional heuristics is not ruled out. In other theories customers use just one method of CSQE.

Sixthly, in the theory suggested here CSQE improvement is step-wise. This means that unless the 'incidence of occurrence’ changes at any of the 7 points of comparison, a customer’s CSQE can only change if his or her attitude towards the service experience or service provider is equated with a different (either higher or lower) point of comparison. Consequntly, perceived service improvement alone does not guarantee CSQE improvement. In other theories CSQE improvement is continuous. Although CSQE improvement is step-wise, movements to successive points of comparison do not necessarily provide equal CSQE improvements. Depending on customers’ comparison expectations, a move from one point of comparison to another may derive large CSQE changes, whilst a move between two other points may derive small CSQE changes. The other theories suggest that equal changes in perceived service lead to equal changes in CSQE.

SUMMARY OF FINDINGS AND MANAGERIAL CONSEQUENCES

When a customer gives a rating or makes a statement about how good a service experience was, or about how good a service provider’s service is, what does he or she mean? How does he or she arrive at those opinions? How can the process which leads to those opinions be efficiently managed? This research has provided improved answers to these questions through multi-disciplinary and empirical research. It has been suggested that the process of service comparison which leads to CSQE is more complex than previously thought, and takes place in a variety of similar ways. Consequently, whilst many avenues of research concerning the above questions remain open, this research has contributed to understanding relating to services marketing. The consequences of this improved understanding for the management of services will be explained in a later article but are believed to be substantial and include more accurate and more detailed consumer focused benchmarking, and improved prediction of the transitional costs and CSQE benefits associated with achieving an improved benchmark position.

REFERENCES

Berry, L.L. and A. Parasuraman (1991), "Marketing Services, Competing Through Quality", Free Press.

Bolfing, C. and R.B. Woodruff (1988), "Effects of Situational Involvement on Consumers’ Use of Standards in Satisfaction/Dissatisfaction Processes", Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behaviour, 1, 16-24.

Boulding, W, Kalra, A, Staelin, and V.A. Zeithaml (1993), "A Dynamic Process Model of Service Quality: From Expectations to Behavioural Intentions", Journal of Marketing Research, 30, 7-27.

Brown, T.J., Churchill G.A.Jr. and J.P. Peter (1993), "Research Note: Improving The Measurement Of Service Quality", Journal of Retailing’, New York University, Vol.69, No.1, Spring, 127-139.

Brown, N.R., Rips, L.J. and S.K. Shevell (1985), "The subjective dates of neural events in very long-term memory", Cognitive Psychology, 17, 139-177.

Cadotte, E.R., Woodruff, R.B. and R.L. Jenkins (1983), "Norms and Expectations: How Different are the Measures?", in Day, R.L. and H.K. Hunt., (eds.) "International Fare in Consumer Satisfaction and Complaining Behaviour", Bloomington, IN, Indianna University, 49-56.

Cadotte, E.R., Woodruff, R.B. and R.L. Jenkins (1987), "Expectations and Norms in Models of Consumer Satisfaction", Journal of Marketing Research, no.24, 305-314.

Churchill, G.A.Jnr. and C. Surprenant (1982), "An Investigation into the Determinants of Customer Satisfaction", Journal of Marketing Research, no.19, 491-504.

Gabrielcik, A. and R.H. Fazio (1984), "Priming and frequency estimation: A strict test of the availability heuristic", Personality and Social Psychology Buletin, 10(1) 85-89.

Goldstein, E.B. (1989), "Sensation and perception", third edition, Wandsworth.

Gregory, R.L (1986), "The psychology of seeing, Third edition: revised and updated", Weidenfeld & Nicolson.

Gr÷nroos, C. (1982), "Strategic Management & Marketing In The Service Sector" Research Report No.8, Swedish School of Economics & Business Administration, Helsinki, Finland.

Kintsch, W. (1980), "Semantic memory: A tutorial", in Attention and Performance, 8, Hillsdale, N.J, Lawrence Erlbaum Associates Inc., 595-620.

Kohonen, T. (1984), "Self organization and associative memory", Berlin, Springer-Verlag.

Lewis, P.E. (1986), "Understanding the process of Customer Service Quality Evaluation: An extension through psychology and empirical study." in Edvardsson B., Brown W.W., Johnston R. and Scheuing E.E., (eds.), "Advancing Service Quality: a global perspective", ISQA.

Liljander, V (1995), "Comparison Standards in Perceived Service Quality", publication of the Swedish School of Economics and Business Administration, Helsinki, Finland, No.63.

Liljander, V. and T. Strandvik (1993), "Different Comparison Standards as Determinants of Service Quality", Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behaviour, 6, 118-132.

Parasuraman, A, Berry, B. and V.A Zeithaml (1990), "An Empirical Examination of Relationships in an Extended Service Quality Model", Working Paper, Marketing Science Institute.

Parasuraman, A, Berry, B. and V.A Zeithaml (1991), "Refinement and Reassessment of the Servqual Scale", Journal of Retailing, Vol.67, no.4, 420-445.

Parasuraman A., Zeithaml V.A. and L.L. Berry (1985), "A Conceptual Model of Service Quality & Its Implications For Future Research", Journal of Marketing, No 49, 41-50.

Parasuraman A., Zeithaml V.A. and L.L. Berry (1988), "Servqual: A Multiple Item Scale for Measuring Consumer Perceptions of Service Quality", Journal of Retailing, Spring, 12-40.

Prakash V. and J.W. Lounsbury (1984), "The Role of Expectations in the Determination of Consumer Satisfaction", Journal of the Academy of Marketing Science, no.12, 63-76.

Rumelhart, D.E. (1975), "Notes on a schema for stories", in Bobrow D.E. and Collins A., (eds), Representation and Understanding, New York, Academic Press, 211-236.

Rumelhart, D.E. (1980), "The building blocks of cognition", In Spiro, R., Bruce, B. and Brewer, W., (eds), Theoretical issues in reading comprehension, Hillsdale, N.J, Lawrence Erlbaum Associates Inc.

Rumelhart, D.E. and D.A. Norman (1985), "Representation of Knowledge", In Aitkenhead, A.M. and Slack, J.M., (eds), Issues in Cognitive Modeling, Lawrence Erlbaum Associates Ltd, 15-62.

Swan, J.E. (1988), "Consumer Satisfaction Related to Disconfirmation of Expectations and Product Performance", Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behaviour", 1, 40-47.

Tversky A. and D. Kahneman (1973), "Availability: A heuristic for judging frequency and probability", Cognitive Psychology, 5, 207-232.

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