The Effect of Conditions of Administration on Listening Comprehension Test Scores

ABSTRACT What constitutes “standard conditions” of administration for tests of listening comprehension? Is it sufficient to present the test through a tape recording which insures a standardized rate of presentation, pronunciation, and sound quality? Or do differences in local pronunciation, in acoustical properties of rooms, and in the extent to which individuals use visual clues while listening suggest that presentation by a familiar individual or by sound motion picture is more desirable? This study was designed to explore these problems. The closed circuit television installation at The Pennsylvania State University presented an opportunity to vary the media of presentation. The presence of comparable student groups permitted a test of the hypothesis that variability among readers of a script resulted in greater differences in performance than variability in media of presentation. Analysis of covariance techniques were used, controlling on scholastic aptitude and on pretest performance on a listening comprehension test administered over TV. Significant differences were found across readers. No significant differences were found across media. There is some evidence that under conditions less favorable than those of this experiment, differences across media might be obtained.