[Previous] | [Session 80] | [Next]
G. L. Deming (Univ. of Maryland)
During 2000 and 2001, the validity and reliability of the Astronomy Diagnostic Test Version 2.0 (ADT 2.0) were formally investigated through the Astronomy Diagnostic Test National Project. The ADT 2.0 was administered as a pre-test to 5346 students and as a post-test to 3842 students. Student test results were collected from 97 classes that ranged in size from 4 to 320 students with 30 states represented. The 68 professors participating in the ADT National Project taught classes at universities (54%), 4-year colleges (27%), and 2-year colleges (19%). The database was analyzed for reliability at the Ontario Institute for Studies in Education. A pre-test value for Cronbach's alpha of 0.65 and post-test value of 0.76 demonstrate an acceptable degree of internal consistency. The average score for the 44 participating professors who completed the ADT as experts was 98%. Face and content validity were established by combining results from the experts with feedback from 60 student interviews. Student results from the National Project yielded an average score of 32.4% for the pre-test and 47.3% for the post-test. There is a gender discrepancy in favor of males that persists in both the pre-test (11% points) and the post-test (12% points) scores. The variations across geographic distribution and institution types were not significant. In addition to the 21 content items, the ADT 2.0 has 12 student background questions enabling instructors to have a better understanding of who takes introductory astronomy.
This research was supported by the National Science Foundation through grants REC-0089239 (GD) and DGE-9714489 (BH).
The author(s) of this abstract have provided an email address for comments about the abstract: deming@astro.umd.edu