The Effect of Scale Form Choice on Psychometric Properties of Patient Satisfaction Measurement

One hundred and fifty nine subjects responded to one of three differently labeled 5-point scale forms of patient satisfaction measurement: Likert-type scale form (1 = Strongly Agree, 5 = Strongly Disagree); Mixed scale form (1 = Very Satisfied, 5 = Very Dissatisfied); and Expectational scale form (1 = More Than I Expected, 5 = Less Than I Expected). A well-known patient satisfaction instrument developed by Ware and colleagues (Ware, Snyder, and Wright, 1976) was utilized to manipulate the scale form effect. The psychometric properties of each scale form was analyzed. The findings indicated generally comparable satisfaction data generated by each scale form as well as comparable measures of reliability. The Likert-type scale form has shown advantage in reducing leniency (relatively high satisfaction scores) over the mixed and expectational scale forms.