Thesis (Ed.D.)--Boston University / One goal of science education reform is student attainment of scientific literacy. Therefore, it is imperative for science educators to identify its salient elements. A dimension of scientific literacy that warrants careful consideration is scientific thinking and effective ways to foster scientific thinking among students. This study examined the use of mental models as evidence of scientific thinking in the context of two instructional approaches, transmissional and constructivist. Types of mental models, frequency of explanative information, and scores on problem solving transfer questions were measured and compared among subjects in each instructional context. METHODS: Subjects consisted of sophomore biology students enrolled in general biology courses at three public high schools. The Group Assessment of Logical Thinking instrument was used to identify two equivalent groups with anN of 65. Each group was taught the molecular basis of sickle cell anemia and the principles of hemoglobin gel electrophoresis using one of the two instructional approaches at their schools during five instructional periods over the course of one week. Laboratory equipment and materials were provided by Boston University School of Medicine's MobileLab program. Following the instructional periods, each subject was asked to think aloud while responding to four problem solving transfer questions. Each response was audiotaped and videotaped. The interviews were transcribed and coded to identify types of mental models and explanative information. Subjects' answers to the problem solving transfer questions were scored using a rubric. RESULTS: Students taught in a constructivist context tended to use more complete mental models than students taught in a transmissional context. Fifty two percent of constructivist subjects and forty four percent of transmissional subjects demonstrated evidence of relevant mental models. Overall fifty two percent of the subjects expressed naive mental models with respect to content. There was no significant difference in the frequency of explanative information expressed by either group. Both groups scored poorly on the problem solving transfer problems. The average score for the constructivist group was 30% and the average score for the transmissional group was 34%. A significant correlation was found between the frequency of explanative information and scores on the problem-solving transfer questions, r = 0.766. CONCLUSION: The subjects exhibited difficulty in formulating and applying mental models to effectively answer problem solving transfer questions regardless of the context in which the subjects were taught. The results call into question the extent to which students have been taught to use mental models and more generally, the extent to which their prior academic experience has encouraged them to develop an awareness of scientific thinking skills. Implications of the study suggest further consideration of mental modeling in science education reform and the deliberate integration of an awareness of scientific thinking skills in the development of science curricula.