This study investigated the effects of three aspects of language testing (test task, familiarity with an interviewer, and test method) on both tester and tested. Data were drawn from several previous studies by the researcher. Concerning test task, data were analyzed for the type of topic students wanted most to talk about or preferred not to talk about, and whether they had similar preferences for Japanese and English tests. Concerning the interviewer factor, data were analyzed for whether the interviewer was a classroom teacher, whether teacher and interviewer could share a common conversation topic, and whether the interviewers were interested in topics the students respond to. Student preferences for oral test method, direct or semi-direct and type of interaction used to elicit speech, were also analyzed. Results indicate that at different proficiency levels, students perform differently on direct and semi-direct tests, and interviewers' choice of test questions influenced student performances and may have even influenced raters' ratings. Implications for fairness in testing are considered. Contains 18 references. (MSE)