Conceptual characterization of threshold concepts in student explanations of evolution by natural selection and effects of item context

Research output: Contribution to journalJournal articleResearchpeer-review

Evolutionary theory explains a wide range of biological phenomena. Proper under­standing of evolutionary mechanisms such as natural selection is therefore an essential goal for biology education. Unfortunately, natural selection has time and again proven difficult to teach and learn, and students’ resulting understanding is often characterized by misconceptions. Previous research has often focused on the importance of certain key concepts such as variation, differential survival, and change in population. However, so-called threshold concepts (randomness, probability, spatial scale, and temporal scales) have also been suggested to be important for understanding of natural selection, but there is currently limited knowledge about how students use these concepts. We sought to address this lack of knowledge by collecting responses to three different natural selection items from 247 university students from Sweden and Germany. Content analysis (deduc­tive and inductive coding) and subsequent statistical analysis of their responses showed that they overall use some spatial scale indicators, such as individuals and populations, but less often randomness or probability in their explanations. However, frequencies of use of threshold concepts were affected by the item context (e.g., the biological taxa and trait gain or loss). The results suggest that the impact of threshold concepts, especially randomness and probability, on natural selection understanding should be further explored.

Original languageEnglish
Article numberar1
JournalCBE Life Sciences Education
Volume19
Issue number1
DOIs
Publication statusPublished - 1 Mar 2020
Externally publishedYes

Bibliographical note

Funding Information:
We thank the whole EvoVis group (Gustav Bohlin, Ute Harms, Gunnar Höst, Nalle Jonsson, Marta Koc-Januchta, Konrad Schönborn, and Jörgen Stenlund) for valuable support during work on the article. Special thanks are due to Gunnar Höst for valuable discussions on statistics and John Blackwell for language review. We also thank the Swedish Research Council for providing the funds for the research presented in the paper. This work was supported by the Swedish Research Council (VR 2012:5344, LT).

Publisher Copyright:
© 2020 A. Göransson et al.

ID: 375592775