Student-Centered Reliability, Concurrent Validity and Instructional Sensitivity in Scoring of Students' Concept Maps in a University Science Laboratory *

Student-centered approach of scoring the concept maps consisted of three elements namely symbol system, individual portfolio and scoring scheme. We scored student- constructed concept maps based on 5 concept map criteria: validity of concepts, adequacy of propositions, significance of cross-links, relevancy of examples, and interconnectedness. With respect to the concurrent validity of scoring the concept maps in this study, correlation coefficients were computed between total scores of students' concept map and their scores of Prior Knowledge Test (PKT), and Achievement Test (AT) involved in the concepts of science laboratory experiments. The mean scores of students' pre-lab concept maps correlated much better with their scores of PKT (r= .615, p < 0.01), and also mean scores of students' post-lab concept maps correlated with their scores of Achievement Test (r= .478, p < 0.05). In regard to instructional sensitivity, the statistical analysis based on the comparison of both total and interconnectedness scores of pre- and post-lab concept maps indicated that there were significant differences favoring the post-lab concept maps (p < 0.01).