Students explain evolution by natural selection differently for humans versus nonhuman animals

Abstract

Evolution is foundational to understanding biology, yet learners at all stages have incomplete and incorrect ideas that persist beyond graduation. Contextual features of prompts (e.g., taxon of organism, acquisition vs. loss of traits, etc.) have been shown to influence both the learning process and the ideas students express in explanations of evolutionary processes. In this study, we compare students' explanations of natural selection for humans versus a nonhuman animal (cheetah) at different times during biology instruction. We found "taxon" to be a significant predictor of the content of students' explanations. Responses to "cheetah" prompts contained a larger number and diversity of key concepts (e.g., variation, heritability, differential reproduction) and fewer naïve ideas (e.g., need, adapt) when compared with responses to an isomorphic prompt containing "human" as the organism. Overall, instruction increased the prevalence of key concepts, reduced naïve ideas, and caused a modest reduction in differences due to taxon. Our findings suggest that the students are reasoning differently about evolutionary processes in humans as compared with nonhuman animals, and that targeted instruction may both increase students' facility with key concepts while reducing their susceptibility to contextual influences.

FIGURE 1.

Prompts used in the HCA . (A) Two forms of an assessment were developed that differed in trait type (structural vs. functional). (B) Each form prompted students (n = 91, Form 1; n = 69, Form 2) to explain evolution by natural selection for both human and nonhuman animals. Students responded to the same form at the beginning and at the end of the semester.

FIGURE 2.

Percentage of responses that contain each of the six key concepts and three naïve ideas pre- and post-instruction. KCs occurred more frequently in cheetah responses and responses written at the end of the semester. NIs occurred less frequently at the end of the semester. For numeric data equivalents refer to Supplemental Table S1.

FIGURE 3.

Average number of KCs in responses for each of the two taxa, estimated by the fitted model.

FIGURE 4.

Average number of NIs in responses for each of the two traits, estimated by the fitted model.

FIGURE 5.

Frequencies of (A) KCs and (B) NIs in student responses pre- and post-instruction.

FIGURE 6.

Average number of (A) KCs and (B) NIs in responses for pre- and post-instruction, estimated by the fitted model.

FIGURE 7.

Changes in the contents of the responses for students’ responses to the (A) cheetah prompt and (B) human prompt. A total of 160 students provided four responses each: one to the cheetah prompt and one to the human prompt at the start of the semester (pre), and the same at the end of the semester (post). Plots created using SankeyMATIC.