METRICAL CHARACTERISTICS OF THE QUESTIONNAIRE APPLIED TO THE QUALITY ASSESSMENT OF THE SERVICES OF HIGHER EDUCATION IN THE REPUBLIC OF SRPSKA
Abstract: In order to validate the research instrument used to identify student perceptions of the quality of service provided by higher education institutions in the Republika Srpska- (Pan-European Apeiron University in Banja Luka, University of Banja Luka, University of East Sarajevo, Prijedor College of Medicine and University of Business Studies Banja Luka) (within a broader empirical study), its metric characteristics were determined. In our study, respondents expressed their estimation by choosing one of the positions on a fi ve-point Likert scale. Th e numerical values of the ordinal scale arranged in this way corresponded to the intensity of the assessment of thirty service indicators, namely: 1 (much less than my expectations); 2 (less than my expectations); 3 (according to my expectations); 4 (more than my expectations); 5 (signifi cantly more than my expectations). Th e value evaluation of the response is interpreted in accordance with the rule: (a) the perception of the quality of service “as expected” (if the scalar average is close to the quantitative value of three), (b) the perception of the quality of the service “below expectations” (when the scalar average is well below quantitative values of three), (c) perception of service quality “above expectations” (if it is a scalar average well above quantitative value of three) (Klarić & Kulišan, 2011). Also, in order to rank the importance of indicators, in manifest space, values of scalar averages (Mean) are used in interpretation (Nešić et al., 2017).Th e questionnaire, based on the HEDUQUAL scale, was tested using two procedures: (1) factor analysis (Principal Components Analysis, Direct Oblimin) and (2) Scale Reliability Analysis - based on Cronbach’s alpha coeffi cient). Th e sample of respondents consisted of a population of students attending one of the higher education institutions in the Republika Srpska, a total of 863 students (M = 265; F = 598) with an average age of 21.5 years. Th e results obtained confi rm the good metric of the scale, both in terms of validity (KMO = 0.972; Bartlett’s Test of Sphericity = 14292.613; Sig. = 0.000) and in the context of its reliability (Cronbach’s Alpha = 0.966). All statistical inferences were conducted at a signifi cance level of 0.05 (Sig. <, 05). Th e good metric characteristics of the instrument defi ned the quality basis for further analytical observations of the study results. Based on good metric characteristics obtained in this part of the study (whose task was to validate the survey questionnaire), it is reasonable to expect that, in continuing the analytical examination of the empirical results of the study, the research conclusions can be reasonably based on accurate and academically correct statistical background. Which is a prerequisite for the correct dissemination of the information obtained and, accordingly, a scientifi cally correct interpretation of the overall results of the study.
Keywords: quality of services; rating scale, metric characteristics.