Consistency of Yes-Sayer behavior across a range of questions

Objectives: The primary goal was to determine whether bogus-distractor identified ‘yes-sayers' consistently demonstrated that tendency to be a ‘yes-sayer' across other factual questions in a questionnaire. “Yes-saying' is defined as giving responses which tend to ‘please the interviewer' or ‘indicate knowledge not truly held'. Methods: These data came from the Tuskegee Legacy Project (TLP) Questionnaire, as administered to 976 adults (353 Blacks and 623 Whites) in four U.S. cities. ‘Yes-sayers' were identified via a subset of 8 ‘have-you-heard-of' questions, comprised of 4 ‘bogus studies' and 4 ‘real studies'. The individuals who responded ‘yes' to having heard of 2 or more of the bogus studies were identified as ‘yes-sayers' (YS), all others were categorized as ‘straight-shooters' (SS). The percentage of YS vs SS, as based upon those 8 questions, was then also calculated for 9 other factual questions on the TLP Questionnaire to determine if this was a consistent behavior. Results: Blacks had a higher percentage of ‘yes-sayers' than did Whites (42.8% vs 24.2%, p <0.05) as identified using the designed subset of 8 questions. However, when checking for consistency as a ‘yes-sayer' across the 9 other factual questions, there were neither sizable nor statistically significant differences, nor a discernible pattern, in % Yes-Sayers within Black or White respondents. Conclusions: Although Blacks were identified as more likely than Whites to be ‘yes-sayers' on a question subset specifically designed to detect that trait via the use of ‘bogus-distractors', there was no evidence that this was a ‘generalized behavioral trait' throughout the questionnaire for either Blacks or Whites. Given that these 9 other factual questions in the TLP Questionnaire were not initially, nor specifically, designed to confirm the behavioral trait of ‘yes-saying', future questionnaire design methodological studies should be conducted using specifically designed ‘consistency-confirmation' questions to check the validity of these findings.

This study was supported by NIDCR/NIH grant U54 DE14257 and T32 DE007255.