We outline three educational implications of Conviction Narrative Theory (CNT).

First, CNT can contribute to understanding pre-existing results in the education literature, especially regarding challenges to effective learning that is retained and applied in new contexts (e.g., Barnett & Ceci, Reference Barnett and Ceci2002). As a concrete example, imagine a biology teacher who hopes that class content will help students understand the world and make better decisions in the future. Perhaps, years after the class, students might be able to correctly identify a new conspiracy theory around vaccines as being biologically implausible. Such an achievement requires that relevant information is initially understood and encoded into memory, remembered over a long time, recalled unprompted and outside of the original learning context, and finally applied correctly to a novel context (“far transfer”). CNT makes clear some of the challenge: When people are constructing potential narratives, snippets of information learned for a very narrow goal and applied in a very narrow context long ago (e.g., memorizing what “Helper T cells” are, to pass a biology exam about the immune system) may have very low salience as building blocks for narratives.

Second, CNT has implications for educational practice: Teachers can help students identify when intuitive narrative is or is not best suited to understanding and decision-making. When narrative is useful, teachers can help students learn to use it more effectively.

Consider an example of narratives interfering with reasoning, such that increased knowledge paradoxically increases wrong judgments. Kahan et al. (Reference Kahan, Peters, Wittlin, Slovic, Ouellette, Braman and Mandel2012) found that “Members of the public with the highest degrees of science literacy and technical reasoning capacity… were the ones among whom cultural polarization was greatest” (e.g., about climate change). Such results indicate the power of the “narratives”: Teaching and learning more “facts” can be useless – or even harmful – if incorporated into a flawed narrative framework. This is consistent with research on confirmation bias (Nickerson, Reference Nickerson1998), but CNT makes clear what is being confirmed in many of these cases: A particular narrative about the world. Regarding education, it suggests that students might be guided to be particularly cautious about narratives when they feel strongly about a topic but don't yet know a lot about it.

In contrast, other examples show narratives being an asset for learning and reasoning. Dudley and Bold (Reference Dudley and Bold1996) revised a non-calculus-based introductory physics course to start with everyday life (e.g., how does electricity within a house work?) and then worked down to the low-level physics phenomena (e.g., voltage). Notably, this approach “does not insist on mathematical rigor at every step, but concentrates more on the development of conceptual understanding.”

These contrasting examples illustrate the difficulty of knowing when to employ narrative. Indeed, some colleagues objected to the revised physics course because “they fear that any neglect of rigor compromises the physics education of the students.” This gap between what instructors think is best and what works well for students may be partially explained by “the curse of knowledge”: Once you know something, it is hard to simulate the mind of someone who does not know it (Wieman, Reference Wieman2007). Imagine a psychology teacher explaining the proper interpretation of p-values. Even when the lesson focuses on the mathematical level, the teacher's rich understanding of p-values is embedded in knowledge of the “replication crisis” in psychology. She might be thinking about Simmons, Nelson, and Simonsohn (Reference Simmons, Nelson and Simonsohn2016) reporting that “undisclosed flexibility in data collection and analysis allows presenting anything as significant”; or Bennett, Miller, and Wolford (Reference Bennett, Miller and Wolford2009) finding “results” from an fMRI of a dead salmon; or check-lists to avoid p-hacking (Wicherts et al., Reference Wicherts, Veldkamp, Augusteijn, Bakker, Van Aert and Van Assen2016); or “meta-discussion” about the tone around discussions of replication (Derksen & Field, Reference Derksen and Field2022). In short, the instructor has a huge amount of related knowledge available to contextualize – and narrativize – her understanding of the low-level details being discussed; the students do not, and this gap can make pedagogy extremely challenging.

Third, CNT suggests priorities for future education research. Future research can identify both how to best use narratives and how to best teach students to use narratives. Consider simulation games that place students in the heart of the action. Shared experience of students in role-playing exercises becomes a useful narrative that the instructor can leverage to translate knowledge to terms meaningful to students. For example, one of us has had students do a classroom exercise simulating the working of direct democracy, with students taking roles with different interests and knowledge in a self-governing democratic city facing a momentous decision. By pursuing their interests in the exercise, students end up experiencing a variety of important social and political phenomena, such as the spread of misinformation. Class comes “built-in” with the narrative of what happened in the exercise that led to the spread of misinformation, providing the instructor with a concrete narrative for analysis to ensure a shared understanding of the concepts and theories. Such practices seem well-supported by various aspects of CNT, but future research can help establish which practices actually lead to the best learning outcomes.

As an additional consideration, how can teachers give clear feedback when the focus is not on whether the final answer is “correct” but rather the process (and narrative) leading to the answer? Teachers must challenge students to break apart and critically examine the narratives underlying decisions and perspectives. This approach presents an opportunity to teach a transferable skill rather than discipline-specific knowledge. Through an evaluative lens: Are the students presenting structured evidence to justify their position? Is this evidence of high source quality? Do valid causal links tie narrative elements? Have they identified and mitigated biases that could have undue influence? Such questions prompt faculty to craft feedback aligned with Hattie and Timperley's task, process, and self-regulatory levels (Reference Hattie and Timperley2007) that is likely to serve as “high-information feedback,” shaping learner behavior in more productive ways than simple reinforcement or punishment (Wisniewski, Zierer, & Hattie, Reference Wisniewski, Zierer and Hattie2020). This deeper engagement with process can enable students to regularly sense-check their biases and assumptions, be more open to incorporating new evidence and updating perspectives, and be more resistant to superficially compelling yet unfounded narratives.