IJMR Landmark Paper: 'Measuring service quality: A review and critique of research using SERVQUAL'
31 October 2018

Measuring service quality: A review and critique of research using SERVQUAL, Lisa J Morrison Coulthard, University of Leicester, IJMR Vol. 46 Q4.

Measuring service quality is a re-occurring theme in submissions to IJMR, many of which either refer to SERVQUAL, or use it as a key measure. As many of these submissions aim to measure specific situations within a limited geographic area, they tend to be rejected as they add little new knowledge to the extensive literature in this field. SERVQUAL has a long history, first described back in 1988, but it has survived as a measurement methodology, despite other approaches, including within papers published in IJMR, such as those by Phil Klaus and Stan Maklan (‘Customer experience: are we measuring the right things?’, IJMR 53/6 and ‘Towards a better measure of customer experience’, IJMR 55/2), who advocate the use of a more holistic method that covers the whole customer experience, rather than either customer satisfaction or service quality. In an earlier IJMR paper by Meng et al (‘Some retail service quality expectations of Chinese shoppers’, IJMR 51/6), the authors discuss the dangers of applying generic models, such as SERVQUAL, developed in western economies, to measure customer satisfaction in emerging economies, that without modification do not adequately address the subtle cultural differences that exist in countries such as China.

I’ve therefore selected Coulthard’s paper as  the latest Landmark Paper as the author provides an in-depth examination of SERVQUAL, based on reviewing the structure of the measures used within the model as described and discussed in reviews of the method. In doing so, the author covers another issue that bedevils many IJMR submissions – the unquestioned use of Likert scales as a method of measurement, without any discussion of their suitability for the particular topic being researched.

Coulthard describes the evolution of SERVQUAL, where quality is treated as a type of attitude, perceived quality being defined by Zeitaml in 1987 as ‘the consumers judgment about an entity’s overall excellence or superiority’.  This evolved into comparing what companies are perceived to offer with what they should offer. From this a gap-based model was developed where perceived survey quality is defined as ‘the degree and direction of the discrepancy between consumers’ perceptions and expectations’, leading to the calculation of positive and negative ‘disconfirmation’.

As Coulthard describes, the original model was revised based on the considerable evaluation and criticism that followed on from the initial 1988 paper. However, replication studies continued to undermine the assertions made for the method, but this could be due to modifications made by others to fit particular sectors. More fundamental were criticisms that the substantial body of knowledge on the psychology of perception was ignored; the likely costs of service improvements (which perhaps could have been addressed through conjoint methods, although evidence in the paper indicates this presents challenges, but the importance of rating service attributes and ways to weight service components have been widely discussed ) and data collected via ordinal methods analysed via methods suited to interval data (a lesser criticism). Other issues about the nature of service quality processes were also raised in subsequent reviews, and the research process, including difficulties in administering the batteries of scales, and the subsequent development of the derivative SERVPERF model.  

As I briefly alluded to earlier in this introduction, the paper contains a lengthy section discussing the challenges and issues in designing and administering Likert scales. Although this is in the context of SERVQUAL, the discussion has a much wider application, especially in terms of the cited sources in this field. The points raised in this section are seldom discussed in submissions to IJMR where Likert scales form the basis of the data collection. There is often an implied suggestion that scales developed by others can be reused, without any consideration as to how they were derived in the first place; whether they are truly appropriate in the new context, and, whether they relate to human cognitive processes and the selected data collection method.

In conclusion, the author notes the popularity of the method, despite the criticisms. The method also evolved at a time when measuring satisfaction and the wider service experience were still in their relative infancy. But, as Coulthard states in the final sentence: ‘An alternative with the same level of general appeal and market dominance is yet to be produced’. Some might argue that the Net Promoter Score is ‘the only number you need’, but this widely applied, and often miss-understood, method also has many critics (personally, I feel that customers deserve more consideration at board level than can be derived from a single measure). Klaus and Maklan advocate a more holistic (EXQ scale) method in their second paper, but the SERVQUAL based submissions still keep flowing in. 

Finally, Parasuraman et al in their original paper (referenced by Coulthard) included a graphical representation of the service delivery process to help pin-point where disconnects can occur that lead to service delivery failing to meet the needs of customers. This also includes the potential gap between how the service is promoted and the perceptions and experiences of customers. Whilst I’ve never applied the SERVQUAL model in studies I’ve developed for measuring customer experience and service quality, I found this process gap analysis a very useful tool for identifying where these disconnects might occur, and where targeted internal research within the organisation, or external with customers, could be very helpful in developing effective solutions to improve the customer experience.

Read the original paper

If you are an MRS Certified Member you can access this paper free of charge. Login to our website, then follow this link to the paper at SAGE Journals.