Type of Document Dissertation Author Craig, Susannah Fowler URN etd-10302007-090619 Title Comparison of Middle School Student Performance on Photograph-Based Multiple-Choice Items and Test-Based Multiple-Choice Items on a State Science Achievement Test Degree Doctor of Philosophy (Ph.D.) Department Educational Theory, Policy, & Practice Advisory Committee
Advisor Name Title James H. Wandersee Committee Chair Earl H. Cheek Committee Member Pam B. Blanchard Committee Member Rita R. Culross Committee Member William B. Stickle Dean's Representative Keywords
- life science test items
- multiple-choice items
- science assessment
- human constructivism
Date of Defense 2007-08-09 Availability unrestricted AbstractThe purpose of this study was to investigate how photograph-based life science multiple-choice items influenced Louisiana science students' performance on statewide standardized tests, in comparison with text-based items about the same content. This mixed methodology research study focused primarily on types of multiple-choice items, specifically five matched pairs of multiple-choice items, text-only and same-text with a photograph. For the 2007 LEAP field test, statistics from 11 multiple-choice items were utilized to characterize student performance on photograph-based multiple-choice items. Data from all Louisiana 8th grade students taking Form 3 (n=1130) and Form 4 (n=1182) were analyzed to compare student performance on each item type.
Additional case study research was conducted in two schools. Within each school, one 8th grade class was exposed to the 20-Question Model (Wandersee, 2000) (treatment group); the remaining 8th grade classes were not (control group). Questionnaires were given to all 8th grade students at each school which focused on the student's experience when answering the field test questions with a photograph. In addition four eighth-grade students, who were contrasted on gender and on high or low academic performance, were interviewed and asked to co-construct six concept maps related to six different test items used in the study (four with photographs, two without photographs).
The analysis of the quantitative data showed a significant difference on the heron item. There was a moderate positive correlation between achievement level and mean number correct on the photograph-based items (rs=.1536). The data show that students performing at low achievement levels benefited from the photograph-based item. The qualitative data analysis revealed positive student perception when working with photographs during classroom instruction and taking assessments. The student interviews and concept maps with the four students revealed students' conceptions and misconceptions about life science concepts.
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access Craig_dis.pdf 2.42 Mb 00:11:12 00:05:45 00:05:02 00:02:31 00:00:12
If you have questions or technical problems, please Contact LSU-ETD Support.