ELECTRONIC STUDENT RESPONSE SYSTEMS AND PEER INSTRUCTION: WHO GAINS? WHO LOSES?

Archived electronic student response system data were analyzed to determine the impact of using peer instruction in large earth science classes for non-majors. When student conceptual question responses were averaged over large portions of the course and binned into nearly equal size groups by success level (high-, medium-, and low-scoring on in-class conceptual questions), significant differences in response patterns were apparent. Students scoring in the higher groups consistently outperformed the low-scoring group by 10-15% (n=400+; p<0.001) on summative assessments; comparisons to the medium-scoring group were mixed. There were no consistently reproducible differences in conceptest question response patterns attributable to age or gender. Students who scored highest on the pre-course geoscience concept inventory were more likely to score higher on in-class conceptual questions, but no obvious correlations were made regarding students who scored lower on that pre-course assessment. The link between course success and attendance is well known and was reproduced in these analyses. Students who did not attend consistently or attended and did not answer all or most questions scored more than 20% lower on summative assessments than those who engaged more frequently with the peer learning approach (p<0.001). More concerning was the preliminary finding that this approach may place some minority groups at a learning disadvantage. Students self-identifying as ethnic minority (n=54) scored about 10% lower on in-class conceptual questions (p<0.001) than randomly selected non-ethnic students. Additionally, analyses of the data over the course of the semester indicated that students who were less successful at answering questions in class were more likely to miss future classes than their higher performing peers. As such, the pedagogical underpinning of using conceptests with peer instruction to aid in student self-assessment was compromised for those students. Further analyses are needed to determine if those students were disenfranchised by this teaching approach or missed more classes for other reasons.