10 Years of Service Quality Measurement: Reviewing the Use of the Servqual Instrument*
Total Page:16
File Type:pdf, Size:1020Kb
101 10 YEARS OF SERVICE QUALITY MEASUREMENT: REVIEWING THE USE OF THE SERVQUAL INSTRUMENT* Simon Nyeck ESSEC BUSINESS SCHOOL, PARÍS Miguel Morales Riadh Ladhari UNIVERSITÉ LAVAL, QUEBEC, CANADÁ Frank Pons UNIVERSITY OF SAN DIEGO, ESTADOS UNIDOS Resumen En 1988, Parasuraman, Zeithaml y Berry elaboraron un instrumento para medir la calidad del servicio. Desde esa fecha, este instrumento ha sido utilizado en numerosos estudios sobre distintas industrias y en diferentes países, tanto por académicos como por profesio- nales. Sin embargo, a pesar de su amplia difusión, pocos estudios tratan los aspectos de dimensionalidad y validez de esta escala de medición. El presente artículo describe las prácticas observadas con relación a estos aspectos a través del análisis de los estudios que han usado SERVQUAL durante los últimos diez años. A partir de una muestra de 60 trabajos empíricos que usan la escala SERVQUAL, se analiza los principales aspectos de validez tratados por cada autor, empleando una plantilla de análisis adaptada del estudio de Stokes y Miller (1975). Con base en los datos disponibles, el estudio sugiere que la escala desarrollada por Parasuraman, Zeithaml y Berry (1988) no presenta una estructura dimensional estable de cinco factores. Finalmente, el artículo evalúa la influencia de las caraterísticas del diseño de la investigación sobre la confiabilidad de SERVQUAL. Introduction aml and Berry (1985, 1988, 1991, 1994), Grönroos (1983, 1984, 1993) and Eiglier During the last decade, research on serv- and Langeard (1987) emphasize the im- ice marketing focused mainly on the anal- portance of conceptualization and meas- ysis of service quality. Consequently, the urement of the service quality construct. studies conducted by Parasuraman, Zeith- Several researchers in this discipline em- phasize the explanation of the perceived quality by using the SERVQUAL dimen- * Trabajo presentado a la 31ª conferencia de la sions, reproducing, in general, the pro- European Marketing Association (EMAC), lle- vada a cabo en Braga, Portugal, del 28 al 31 de cess followed by Parasuraman et al. mayo de 2002. (1988). Año 7, n.º 13, diciembre de 2002 102 esan-cuadernos de difusión The popularity of SERVQUAL with fulness, and empathy. These five dimen- researchers can be explained mainly by sions are broken into 22 items. Each item its ease of use and by its adaptability to is further split into two more items, one diverse service sectors. Even if certain measuring the expectations having to do researchers have only retained the con- with those companies belonging to the cept of gap analysis as operationalization service sector in question, the other serv- of perceived service quality, it appears ing to measure the perception of the serv- that the SERVQUAL model remains the ice offered by a particular company xyz. most complete attempt to conceptualize and measure service quality. Neverthe- Since the scale was developed using less, over the years, the acceptance of the customers of five service sectors (repair model proposed by PZB as a «standard» and maintenance of small electrical ap- instrument was called into question. Au- pliances, banking, long distance tele- thors thus proposed other conceptualiza- phone, title brokerage, and credit cards), tions (Grönroos 1993; Haywood-Farmer Parasuraman et al. (1988) concluded that 1988; Iacobucci et al. 1994; Johnston «SERVQUAL offers a variety of poten- 1988) as well as other measurement in- tial applications. It can be used to evalu- struments (Cronin and Taylor 1992; ate the expectations of customers... as well Brown et al. 1993). as their perceptions... for a wide range of services and distribution organizations». This study has a twofold objective: Several researchers have therefore used first, the research describes a state of prac- SERVQUAL to measure service quality tices regarding validity in research that in various sectors such as health (Babakus has used SERVQUAL; second, the re- and Mangold 1992; Headley et al. 1993), search evaluates the influence of research banking (Brown et al. 1993; Pitt et al design characteristics on SERVQUAL 1995), fast food (Lee and Ulgado 1997), reliability. professional services (Freeman and Dart 1993), retail trade (Gagliano and Hath- 1. Conceptual Background: The cote 1994), and advertising (Quester et SERVQUAL Instrument al. 1995). For PZB (1988), perceived quality is the Nevertheless, the dimensionality and result of the comparison between what the psychometric properties of SERV- consumers consider the service offered QUAL have caused a lively controversy, by the company (i.e., their expectations) since the studies which have used SERV- and their perceptions of the performance QUAL do not always mention a standard of the service provided. Service quality is dimensional structure. In effect, the sta- thus considered to be the difference be- bility of the factorial structure is not dem- tween the perceptions and the expecta- onstrated, nor is its invariance across var- tions of consumers. In 1988, Parasura- ious sectors proven. This led Babakus man, Zeithaml and Berry broke down ten and Boller (1992) to conclude that the dimensions into 97 items (approximately measure of service quality offers a chal- 10 items per dimension). Five dimensions lenge. Furthermore, the results regarding were finally retained: reliability, presence validity of the instrument are mitigated. of tangible elements, confidence, help- This conclusion concerning the dimen- Año 7, n.º 13, diciembre de 2002 10 Years of service quality measurement 103 sionality and psychometric properties of the effect of the number of points on the the scale also appears in several studies scale on reliability. Bendig (1953, 1954) (Asubonteg et al 1996; Kettinger and Lee observed that reliability is independent 1994; Csipak et al. 1994) which provide of the number of points on the scale. Fur- a comparative evaluation of works hav- ther, Churchill and Peter (1984) and ing used SERVQUAL. Peterson (1994) found significant rela- tionships between reliability and the 2. Research Methodology and Data number of points on the scale. We expect a positive relationship between these two This research sets out with a dual pur- variables. pose: First, we describe research practic- es regarding the utilization of SERV- P3. The number of items: The results QUAL. We pay particular attention to obtained by Churchill and Peter (1984) the evaluation of reliability, convergent, and by Peterson (1994) show the pres- discriminant and predictive validities, as ence of this positive relationship between well as to the dimensionality of the SERV- the two variables. We expect the alpha QUAL instrument; second, this research coefficient will increase as a function of evaluates the existing relationship be- the number of items. tween the research design criteria and SERVQUAL’s reliability. We must men- The Sample tion, however, that our research will be limited to the study of the effect of de- The sample is made up of forty (40) arti- sign on the reliability coefficients because cles published since the appearance of the indicators concerning convergent, dis- SERVQUAL (1988) that we have col- criminant and predictive validities are lected from 18 periodicals. However, an sometimes nonexistent in various studies. article could contain the analysis of serv- ice quality in more than one sector; in The studies carried out by Churchill such a case we considered the sector ex- and Peter (1984) and Peterson (1994) amined to be a sample unit, which gives a have allowed us to select those design total of 60 observations. For an article to criteria that affect reliability. Thus, (1) be retained, it had to fit the following the sample size, (2) the number of points three criteria: use of SERVQUAL or of on the scale, (3) the number of items. a modified SERVQUAL scale; study of The following research propositions are service quality in a given sector follow- mentioned: ing an empirical method; and supplying, in results, indicators concerning the reli- P1. Sample size: Churchill and Peter ability, validity or dimensionality of the (1984) found a negative relationship be- scale (Table 1). tween sample size and the alpha coeffi- cient. However, Peter (1994) observed Data Analysis Grid the absence of such a relationship. The research used Stokes and Miller’s P2. The number of points on the scale grid (1975) to evaluate the use of SERV- used: The available literature does not QUAL scale. The evaluation grid was seem to show any agreement regarding divided into four headings: General char- Año 7, n.º 13, diciembre de 2002 104 esan-cuadernos de difusión Table 1: Description of the sample of the evaluated studies Number of Number of Publication articles sectors inventoried inventoried (n) Asac proceedings 3 3 Advances in Consumer Research 1 1 Decision Sciences 1 1 International Journal of Service Industry Management 4 5 Journal of Business Industrial Marketing 1 1 Journal of Business Research 1 1 Journal of Health Care Marketing 4 5 Journal of Marketing 1 2 Journal of Professional Services 2 3 Journal of Retailing 7 17 Journal of Services Marketing 4 4 Managing Service Quality 1 2 MIS Quarterly 1 3 MRN 1 1 Public Administration Quarterly 1 1 Service Industries Journal 6 8 Total Quality Management 1 2 Total 40 60 acteristics, formulation of the problem, all cases, we classified the studies by tak- data collection, and analysis of data. The ing into account only the information pro- final part of the evaluation grid contains vided within each article. As for reliabili- six criteria. We determined the number ty, we took Cronbach alpha mean, when of final dimensions presented in each sec- the alphas were given by dimension. tor inventoried. Finally, we categorized the convergent, discriminant and predic- Finally, to complete the explanatory tive validities in three ways: the article part of our research, we recoded the ini- presented statistical validation results; the tial data in order to regroup it into cate- article discussed only the validity with- gories, as was done by Peterson (1994).