A Comparative Study on the Methods for Weighting the Dimensions of Customer Satisfaction with Importance Perceived by Customers

Kang, Myunggeun; Cho, Woohyun; Lee, Sunhee; Choi, Kuison; Mooon, Kitae

Journal of Korean Society Quality Assurance Health Care 2000;7(2): 230.
Published online December 30, 2000.

고객만족도 조사도구의 차원별 가중치 부여방법 비교

강명근¹, 조우현², 이선희¹, 최귀선¹, 문기태²

¹이화여자대학교 의과대학 예방의학교실
²연세대학교 의과대학 예방의학교실

A Comparative Study on the Methods for Weighting the Dimensions of Customer Satisfaction with Importance Perceived by Customers

Myunggeun Kang¹, Woohyun Cho², Sunhee Lee¹, Kuison Choi¹, Kitae Mooon²

¹Department of Preventive Medicine, College of Medicine, Ewha Womans University
²Department of Preventive Medicine, College of Medicine, Yonsei University

Abstract

Background
The measuring instruments for customer satisfaction in hospitals are often composed of some dimensions reflecting the conceptive complexity of them. Then, overall satisfaction would be expected to be equal the 'weighted' sum of scores by dimensions because the importance rated by customers may be different across the dimensions. But the issue of how to weight the dimensions with importance is not yet solved. We examined 3 sets of weighting methods as to make effect on predictive power against overall satisfaction.
Methods
We conducted a survey included 483 subjects who had visited or admitted to a university hospital, using the short form questionnaire being developed by The Korean Society of Quality Assurance in Health Care for out-patient and in-patient. By using a multiple linear regression model, we compared among changes of explanatory powers against overall satisfaction as dependent variable after weighting 4 dimensions of the survey questionnaire as independent variables with importance scores of dimensions perceived by consumers. And we compared the feasibility of each weighting, methods by checking missing cases.
Results
There were no weighting methods increasing the explanatory power after applying them. The method of absolute scoring was found higher explanatory-power than others, but this finding had no statistical significance. Regarding the number of missing value, method of absolutely scoring had the least cases.
Conclusion
Our findings suggested that weighting the dimensions with importance might have little significance in the cases of scales having items highly correlated, such as consumers' satisfaction. Though asking with items to be answered absolutely, customers might be rating relatively in some degree and this method produced least missing cases. Considering these points, in the cases when weighting the dimensions with importance would be required, we suggest that weighting method by absolute scoring might be better than others.

Key words Weighted scale;Customer satisfaction;Importance of dimensions;