Barbara H. Beach

Development and validation of a measure of dental patient satisfaction.

As part of the Washington State Dental Auxiliaries Project, a 42-item measure of patient satisfaction with dental care was developed. The measure is comprised of 13 subscales: dentist–patient relations, technical quality of care, access, patient waiting time, cost, facilities, availability, continuity, pain, auxiliaries performing expanded duties, staff–patient relations, staff technical quality of care, and office atmosphere. The measure was developed from a set of 52 items included in a questionnaire administered to the patients of private dental practices in Washington state. Usable questionnaires were returned by 30.8percnt; of patients receiving questionnaires in 1979, 40.1% in 1980, and 34.0% in 1981. Factor analysis plus categorization of items by a panel of professionals were used initially to group items into subscales. Contribution to internal consistency was the final criterion for an items inclusion in a subscale. Internal consistency of subscales ranged from 0.44 to 0.80. The concurrent validity of subscales was assessed by relating patient satisfaction to characteristics of the dental practices. The following statistically significant relationships between subscales and criterion variables were observed: dentist–patient relations and percent of patients seen by the dentist; access and number of weeks appointments must be booked in advance; patient waiting time and actual patient waiting time; continuity of care and percent of patients seen by the dentist; auxiliaries performing expanded duties and delegation to auxiliaries; and staff technical quality and percent of hygienist restorations with satisfactory quality. Each relationship was in the expected direction.

Five studies of subjective equivalence

Abstract Subjects set intervals around various numerical values to indicate the ranges within which they would regard unaided subjective judgments of those values as being essentially correct. This was done for five different sets of numbers: historical dates, proportions, peoples ages, seriousness of diseases and life events, and the money values of inheritances and gifts. In some cases the interval breadths were determined by the magnitudes of the numbers involved; in other cases they were determined by task characteristics; and in yet other cases they were determined by both magnitudes and task characteristics but in very intuitively reasonable ways. Ties between this research and studies of subjective judgments of variances and of credible intervals, as well as lexicographic choice, are examined and various implications of this line of research are explored.

The METRO study: A closer look at mispredicted decisions☆

Abstract In earlier research it was found that errors in prediction about peoples decisions were predominantly of one kind: people who were predicted to make a positive decision were less likely to do so than were people who were predicted to make a negative decision (about whether to have a child). Examination of the former groups subjective expected utilities showed them to favor the positive decision but only moderately. In the present study it is suggested that these people are “potential behaviors”; that is, under the proper conditions these people should be more likely to decide to perform the predicted way than people who were predicted to not perform the behavior and who, in fact, did not. Here the behavior of interest was commuting by bus (vs car). Using a scheme based on multiattribute utility considerations, potential bus riders were identified. Among other things, transportation costs were found to be very important to them. Free bus tickets were sent to them and to a control group. Significantly more potential riders used the tickets than did members of the control group.