Abstract
The present study examined the basis for the postdiction superiority effect the superior accuracy of retrospective or postdiction judgments of test performance in comparison to prediction judgments that has been shown in metacomprehension of text material. Two experiments were conducted in which participants read 4 narrative texts and then answered 16 questions per text. Predictions and postdiction judgments were elicited in either 16-question set blocks or multiple sets of 4 questions each. A postdiction superiority effect was found in both experiments for the multiple-question set condition when participants made judgments over individual 4-question sets (i.e., a local effect), but was not consistently found when judgments were made over the overall 16question sets (i.e., a global effect). In addition, judgment accuracy did not increase over successive question sets or texts. The findings support a feedback h othesis, in which self-generated feedback from answering questions allows for greater postdiction accuracy. An alternative hypothesis based on increased knowledge of the nature of the tests was not supported.
Pierce, Benton Harold (1998). The postdiction superiority effect in metacomprehension of text. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1998 -THESIS -P54.