An investigation of alternative methods for measuring quality and their relationship to visitor satisfaction in a festival setting

Abstract

A variety of conceptual paradigms and alternative approaches to quality measurement have been reported in the tourism literature. This study tested a revised model which incorporates two levels of quality: desired and adequate. This conceptual framework added to the development and use of seven alternative measures of quality: (1) a direct perceptions format; (2) a direct one-column format; (3) a direct two-column format (containing two alternatives); (4) a differential threecolumn format (containing two alternative measures); and (5) a differential perceptions minus expectations format. The objectives of this study were 1) to identify which alternative approach to measuring quality was most valid, reliable, effective, and efficient, and 2) to identify the relationship between perceptions of quality and satisfaction. To meet the objectives, four instruments were developed, All respondents completed an a priori expectations instrument before entering the site, and responded to the Festival's performance on those same attributes in one of three versions of a mail-back questionnaire. A probability sample of 822 festival visitors were handed a mail-back questionnaire, which 67% of them returned. A series of chi-square tests discounted sub-sample differences. Oblique factor and reliability analyses on each of the seven quality measures generally confirmed the a priori assignment of the festival attributes into four domains. Similar factor and reliability analyses on the satisfaction scale confirmed its appropriateness. The seven alternative measures were evaluated against 15 criteria. The three criteria related to scale friendliness were number of items to be rated, difficulty to complete, and response error. The criteria related to domain structure were measured by reliability coefficients and stability of factor dimensionality. Variation was measured by standard deviations. Performance of the measures was assessed on four types of validity: predictive, convergent, discriminant, and nomological. These results supported the study's initial conceptualization of quality and satisfaction being different but related constructs. Five different criteria were used to assess the diagnostic value and managerial usefulness of the measures. When evaluated against these 15 criteria. the two measures incorporated into the three-column approach were substantially superior to the other five measures. It is more valid and reliable, easier to implement, and likely to give managers considerably more insight into visitors' perceptions of their enterprise's performances, than measures currently used.