Cognitive Level Demands of Test Items in State-Adopted Computer Science Textbooks (open access)

Cognitive Level Demands of Test Items in State-Adopted Computer Science Textbooks

Test items supplied with seven textbooks approved for use in Computer Science I and II curricula in Texas public schools were categorized by Bloom's taxonomy of educational objectives. Eating was done by a panel of ten judges selected from a group of participants at a taxonomy workshop. The selection criterion was demonstration of at least 80 percent competency in item classification. Judges received a small stipend for completing the rating task. Of 2020 possible items, 998 were randomly selected for analysis. Equal percentages of items from each text were then randomly assigned to each rater. All statistical analyses were computed using SPSS/PC+ (version 2.1). In both courses, CLD frequencies decreased through the three lower levels. The percentage of questions falling in these levels was approximately 83 percent for both courses. However, the higher-level course contained almost 10 percent more Knowledge level questions than did the lower course. At the higher taxonomic levels, the decline was roughly five percent per level in CS I but erratic in CS II. Analysis by book also revealed wide differences within each course.
Date: August 1988
Creator: Aman, James R.
System: The UNT Digital Library