Construct validity was evaluated by testing for hypothesized positive associations between trust and patients' satisfaction, physician humanistic
behaviors during the visit, patients' active choice of their physician, duration of the relationship, preference for the physician playing a more active role,
and expectation for proportion of care from the physician. A positive association between trust and satisfaction, as measured by the Medical Interview
Satisfaction Scale, was reported by Anderson and Dedrick.7 The association with the physician humaneness was hypothesized because many of the
behaviors assessed by this scale were reported to increase trust in a previous qualitative study using patient focus groups.6 Based on findings from
psychological studies of interpersonal trust 20,21 it was hypothesized that patients who chose their physician based on personal knowledge would have
greater trust. An association between greater trust and a preference for greater physician control of care follows from the common definition of trust as
"a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another."22
Construct validity was further assessed by examining the association between trust in the physician, as measured by the Trust in Physician Scale, and a
general measure of interpersonal trust, the Trust in People Scale.16 Whereas a positive association was anticipated, it was expected to be weak, based
on the hypothesis that the Trust in Physician Scale measures trust in a specific patient-physician relationship rather than on general interpersonal trust.
Associations were tested using Pearson correlation coefficients, t tests, and analysis of variance