Martin Kaplan discusses the effects of interviewer biases in the processes of information elicitation in this article. The author used participants to test the effects different personalities have on garnering information from others. Interviewers were asked to make behavior predictions to test how accurately they were able to elicit information from a subject.
The author tested two main hypotheses: differences in the interviewer's predictions can be explained by personality groups eliciting different information, and that if each group elicits different information, then the information would be consistent. The experiments would test if actual differences existed and were consistent.
Kaplan conducted three tests to attempt to isolate a bias variable and prevent differences in perception to affect the interviewers judgement when forecasting an individual's behavior. The subjects were divided as 'repressors', 'sensitizers', and 'neutrals'. The subjects were divided based on their scores on the Byrne Scale of Repression-Sensitization test. Twenty random subjects were selected from each group as 'judges' and each judge was paired with a neutral 'target'.
The author then explains the specifics of the testing and instructions given to all participants. He explained how he analyzed each group's discussions and interviews and compares factors like speech time of judges and of targets and how participants rated their own personalities. Kaplan made his own predictions on how he thought the judges predictions would differ.
What Kaplan found is that sensitizers tended to speak more in the interviews than their targets did. A significant difference in relevance to criteria behavior was not found between the groups. Kaplan found that what he expected regarding the differences between behavior predictions: sensitizers were more negative, repressors were more positive.
While this study shows that conclusions about information elicited from a subject can be affected by internal biases, the experiment was limited in scope. It's important to mention that the researcher tried to counter potential external variables from affecting the study because this limited the affect that personal perception had on the information elicited from the subject. This study is also pretty old, but because of the type of analysis it attempts to do it seems to be still highly relevant the methodology.
KAPLAN, M. (1968). Elicitation of information and response biases of repressors, sensitizers, and neutrals in behavior prediction. Journal Of Personality, 36(1), 84-91. doi:10.1111/j.1467-6494.1968.tb01461.x