Learning critical thinking skills online: can precision teaching help?

Abstract

Critical thinking is identified as a key educational outcome in higher education curricula; however, it is not trivial to support students in building this multifaceted skill. In this study, we evaluated a brief online learning intervention focusing on informal fallacy identification, a hallmark critical-thinking skill. The intervention used a bite-sized video learning approach, which has been shown to promote student engagement. Video-based learning was implemented within a precision teaching (PT) framework, which modulates the exposure of individual learners to the learning material to enable them to build ‘fluency’ in the targeted skills. In one of the learning conditions, PT was applied synergistically with domain-general problem-based training to support generalisation. The intervention consisted of two learning episodes and was administered to three groups (learning conditions) of 19 participants each: a PT fluency-based training group; a PT + group, where PT was combined with problem-based training; and a self-directed learning control group. All three groups showed comparable improvements in fallacy identification on taught (post-episode tests) and unseen materials (post-intervention assessment), with lower-scoring participants showing higher gains than high-scoring participants. The results of the knowledge retention tests a week later were also comparable between groups. Importantly, in the domain-general fallacy-identification assessment (post-intervention), the two PT groups showed higher improvements than the control group. These findings suggest that the integration of bite-sized video learning technologies with PT can improve students’ critical-thinking skills. Furthermore, PT, on its own or combined with problem-based training, can improve their skill to generalise learning to novel contexts. We discuss the educational implications of our findings.

Liaoning University ChopMelon Net Project: Innovative Research on Sustainable Education Based on Real Social Issues

“…The results of this study showed that the use of ChopMelon Net significantly improved users' critical thinking and problem-solving skills, which is consistent with previous research that suggests that integrating real-world problems into instruction is an effective way to improve these skills [52,53]. The complexity and dynamics of real-world problems require learners to use and integrate knowledge from different domains while encouraging them to adopt an active learning attitude [54].…”

Generative AI in Education and Research: Opportunities, Concerns, and Solutions

“…Presenting students with multiple AI-generated answers to a problem or question and asking them to evaluate the correctness and reasoning behind each solution encourages active engagement with the content and the development of their analytical skills. Another way to facilitate understanding of the problem-solving process is to encourage students to follow the logic employed by AI in reaching a given solution (example in Supporting Information, Section C). , This will help them understand the steps involved and recognize the importance of structured thinking in problem-solving. For example, in a chemistry problem involving stoichiometry, an AI-generated solution might involve identifying the limiting reactant, applying relevant stoichiometric coefficients, and using the given data to calculate the amount of a product formed.…”

“…However, the influx of information necessitated stronger information and data literacy skills. Today, AI-powered systems are altering how we acquire and process information. These tools might reduce the need to memorize basic facts or perform simple reasoning tasks but cannot replace the necessity for high-level educational goals such as analysis, evaluation, or synthesis of new material. , The capacity to synthesize, connect, and apply scientific concepts in novel ways remains paramount. In response to whether instructors should continue to assess students based on writing clarity and formatting, we posit that it is not necessary to grade students solely on their writing skills, particularly when advanced AI tools can accomplish these tasks.…”

“…By analyzing each step of the AI’s approach, students can learn how to deconstruct complex problems into manageable parts and apply relevant principles to solve them. This method improves their problem-solving skills and reinforces their understanding of fundamental concepts. Additionally, educators can use AI-generated solutions to highlight common misconceptions or errors, achieving higher-level educational goals. − As an example, GPT-4 was prompted to answer an exam question previously written by the coauthor for an introductory Physical Chemistry for Life Sciences course. GPT-4 failed to adequately address the prompt (detailed in Supporting Information, Section C) due to its inability to apply the second law of thermodynamics.…”

“…Additionally, educators can use AI-generated solutions to highlight common misconceptions or errors, achieving higher-level educational goals. − As an example, GPT-4 was prompted to answer an exam question previously written by the coauthor for an introductory Physical Chemistry for Life Sciences course. GPT-4 failed to adequately address the prompt (detailed in Supporting Information, Section C) due to its inability to apply the second law of thermodynamics.…”