EFFECTS OF PREEXAMINATION DISCLOSURE OF ESSAY TOPICS

To determine the effects on test performance and test validity of rele asing essay topics before a writing examination, researchers asked 300 prospective graduate students to write 2 essays, 1 on a previously un seen topic and 1 on a topic that was made available about 2 weeks befo re the test. Although for each of the 4 essay topics studied, performa nce was slightly better when a topic had been disclosed, overall analy ses did not reveal any statistically significant effect of disclosure on test performance. Consequently, there was no detectable effect of d isclosure on the relation of test performance to several independent i ndicators of writing skill. Limitations of the study are discussed in terms of its generalizability to a high-stakes testing situation and t he extent to which it addressed both the positive and the negative eff ects of disclosure.