Refining Kano's ‘quality attributes–satisfaction’ model: A moderated regression approach

This project proposes a moderated regression approach that corrects the flaws of the dummy regression method and produces more accurate attribute classification.

Abstract

Service and product quality is a multi-dimensional construct and not all quality attributes are viewed as equally important to customers. Several studies have applied the dummy variable regression method to recognize those product and service attributes that contribute more significantly to customer satisfaction. All attributes are classified into three quality factors (basic, performance, excitement) with each factor having different implications for customer satisfaction. This project proposes a moderated regression approach that corrects the flaws of the dummy regression method and produces more accurate attribute classification. The proposed method was validated using data collected from an online tax declaration service.