UNIVERSITY STUDENTS’ MENTAL MODELS OF THE GREENHOUSE EFFECT: A COMPARISON OF TWO LEARNING ACTIVITIES IN MOVING STUDENTS TOWARD EXPERT THINKING

Student mental models of the greenhouse effect rarely match the mental models of climate scientists. We assessed student thinking throughout a sequence of learning activities in an introductory university-level science course. The purpose was to evaluate what activities changed student thinking, and how their thinking changed. The learning sequence included a common lesson, in which all students participated, followed by one of two contrasting lessons, with students randomly assigned to one or the other. In one of the contrasting lessons, students explored an interactive simulation representing the molecular behavior of greenhouse gases (http://phet.colorado.edu/en/simulation/greenhouse, referred to as the “PhET” lesson); in the other, students explored data showing absorption spectra of different atmospheric gases (the “Data” lesson).

We used concept sketches, multiple choice, and short answer questions as assessments. Concept sketch results are discussed here. Students generated sketches four times: (1) prior to the lesson sequence, (2) after the common lesson, (3) a few days after the last lesson, and (4) on the final exam for the course, 7 weeks later. 164 students completed all the assessments, 76 completed the PhET lesson and 77 completed the Data lesson. 11 students missed these contrasting lessons.

We coded the student sketches and compared how closely they matched “expert” features for 28 coded statements. This preliminary analysis indicates that students’ initial sketches matched “expert” responses for ~50% of the statements, on average. The match improved to ~75% by the final assessment.

Students made the greatest gains toward more expert-like mental models from before to just after the common lesson. When tested a few days after the contrasting lessons, the PhET group performed significantly higher than the Data group though neither group had made significant gains from their performance just after the common lesson. After studying for the final exam, both the PhET and Data groups improved significantly, with no measurable difference between them. Further detailed analysis will show which particular features appeared in student sketches after either lesson, which should yield insight into the efficacy of these lessons on specific aspects of student thinking about the greenhouse effect.