The effects of a summer research training program on science process skills, self-efficacy towards creative productivity and project quality

Date of Completion

January 1992


Education, Special|Education, Secondary|Education, Sciences




The research in science and gifted education suggests that participation in a research training program may affect students' self efficacy towards creative productivity (i.e., the belief in one's ability to perform research), their science process skills, adult productivity, and the subsequent quality of such research. To learn more about research experiences for academically advanced students, this study investigated the relationship between a summer research training program and students' self efficacy towards creative productivity, science process skills and the quality of students' research projects.^ Seventy secondary students who participated in Project Oceanology's summer program were randomly assigned to an experimental or control group. Baseline data included Starko's (1986) Self-efficacy for Creative Productivity Inventory, and Okey and Dillashaw's (1980) Test of Integrated Process Skills (TIPS). The scores of students who received treatment were compared to those who did not. Delayed treatment was provided for the control group after testing was completed. Treatment received by this group was irrelevant to this study. Guidelines established by the International Science and Engineering Fair were used to evaluate the quality of the research produced by the experimental group. Hierarchical regression analyses were used to answer all research questions.^ The analysis of the data revealed no significant difference in self efficacy due to group membership. Of the independent variables used to statistically equate the groups, only academic grades proved to be a significant predictor of self efficacy. In regards to science process skills, a significant but negative effect due to treatment was found. The best predictor for this construct was the students' standardized achievement scores. When the problem due to multicollinearity between variables was controlled for, the significance of the experience variable emerged. Self efficacy did not prove to be a significant predictor of research quality after controlling for variation in academic grades and previous research experience. Previous research experience was a significant predictor of quality. ^