Reading Retention as a Function of Method for Generating Interspersed Questions.

摘要

An experiment with 356 Navy recruits compared the effects on reading retention of adjunct (interspersed) questions generated by four different procedures: human linguistic processing, the AUTOQUEST computer program, and two types of cloze algorithms. Results showed that cloze questions interfered with retention and direct improvement effects from AUTOQUEST were as large as those from human-generated questions. No indirect effects of human or computer questions were observed.