Characterizing Students' Ideas about the Effects of a Mutation in a Noncoding Region of DNA.

Understanding student ideas in large-enrollment biology courses can be challenging, because easy-to-administer multiple-choice questions frequently do not fully capture the diversity of student ideas. As part of the Automated Analysis of Constructed Responses (AACR) project, we designed a question prompting students to describe the possible effects of a mutation in a noncoding region of DNA. We characterized answers from 1127 students enrolled in eight different large-enrollment introductory biology courses at three different institutions over five semesters and generated an analytic scoring system containing three categories of correct ideas and five categories of incorrect ideas. We iteratively developed a computer model for scoring student answers and tested the model before and after implementing an instructional activity designed to help a new set of students explore this concept. After completing a targeted activity and re-answering the question, students showed improvement from preassessment, with 64% of students in incorrect and 67% of students in partially incorrect (mixed) categories shifting to correct ideas only. This question, computer-scoring model, and instructional activity can now be reliably used by other instructors to better understand and characterize student ideas on the effects of mutations outside a gene-coding region.

Group Random Call Can Positively Affect Student In-Class Clicker Discussions.

Understanding how instructional techniques and classroom norms influence in-class student interactions has the potential to positively impact student learning. Many previous studies have shown that students benefit from discussing their ideas with one another in class. In this study of introductory biology students, we explored how using an in-class accountability system might affect the nature of clicker-question discussions. Clicker-question discussions in which student groups were asked to report their ideas voluntarily (volunteer call) were compared with discussions in which student groups were randomly selected to report their ideas (random call). We hypothesized that the higher-accountability condition (random call) would impress upon students the importance of their discussions and thus positively influence how they interacted. Our results suggest that a higher proportion of discussions in the random call condition contained exchanges of reasoning, some forms of questioning, and both on- and off-topic comments compared with discussion in the volunteer call condition. Although group random call does not impact student performance on clicker questions, the positive impact of this instructional approach on exchanges of reasoning and other features suggests it may encourage some types of student interactions that support learning.