Using AI to Bridge the Research-Practice Gap
Plug-and-play prompts for using AI to help access and translate high-quality research.
The world of social-emotional learning is rich with research, with CASEL’s new Social and Emotional Learning: Research, Practice, and Policy journal serving as the latest example of the field expanding its efforts to increase access to peer-reviewed articles focused explicitly on SEL.
However, a multitude of barriers keep most of this valuable research from being applied in educational settings. While educators are clearly in search of research that can be directly applied in the classroom, gaining access to high-quality studies often proves challenging. Even when they locate research that isn't restricted by paywalls, the task of deciphering academic jargon (sorry, researchers) to extract practical insights that can inform their instruction presents its own set of difficulties.
This disconnect is highlighted by a 2019 study series from the Jefferson Education Exchange. The findings show that while teachers favor actionable research that is relevant to their professional duties, a mere 16% of educators actually incorporate research findings into their teaching methods.
It is clear, then, that a central obstacle for educators lies not just in obtaining this rich repository of information but in converting it into practical, actionable methods. This is precisely where AI has the potential to make a significant impact.
Several companies are already designing AI to help make education research more accessible and actionable. But free tools such as ChatGPT already have the potential to sift through vast amounts of research and present it in a more digestible format. AI tools can summarize key findings, suggest actionable steps, and even adapt their responses based on the educator's specific needs.
In this article, we discuss promising approaches for using AI to bridge the gap between research and classroom practice, a subject that I am particularly passionate about. This AI-driven method aims to assist educators in not only accessing high-quality education research as well as translating the findings of that research into practical advice that they can apply to their teaching methods.
Guiding Principles for Research-Practice Integration with AI
There are best practices that we can follow when using AI to bridge the research-practice gap. These guiding principles can help ensure that we maximize the benefits and minimize the risks when using tools like ChatGPT for this purpose.
Prompt for specificity to avoid generic or irrelevant information. AI tools are only as good as the prompts they receive. One of the risks of using AI is the potential for generating generic or irrelevant information. To mitigate this, be specific in your queries. Instead of asking, "Tell me about SEL research," you might say, "Summarize the latest research on SEL strategies for improving student engagement with middle schoolers."
Specify topic-aligned sources for AI to pull from. The credibility of the information you receive is directly related to the sources from which it is derived. When using AI to summarize research, specify the journals, authors, or types of studies you're interested in. For example, you could prompt, "Summarize key findings from the Journal of Educational Psychology on SEL strategies for classroom management."
Use a variety of questions to move from broad findings to specific takeaways. In the sample prompts below, you will start to see how we can ask AI to teach us about a field of research at a high-level (“What does the research say about teaching self-management skills to students?”) as well as recommend specific practices that we can apply to our pedagogy based on the research (“What are some actionable steps that teachers can take to build self-management skills with students based on the latest SEL research?”).
Always verify and cross-reference. While AI can provide quick summaries and insights, it's always a good idea to cross-reference these findings with the original research whenever possible. This ensures that you're getting the most accurate and nuanced understanding of the topic at hand. Specifying sources from AI to pull from is one way to decrease the risk of fake or incorrect findings.
Now, let’s explore how to put these guiding principles into practice.
Example Prompt #1: Research Summary
This prompt can be used with ChatGPT to help educators understand what the research says about specific topics.
You are an experienced researcher and educator with expertise in K-12 education research. Your purpose is to help educators access high-quality education research as well as analyze the findings from that research to write concise, understandable summarizes that educators can learn from.
First, you will ask me for a research topic by saying: “Hi! What research topic are you interested in? Ask me: "what does the research say about..." and I will generate a concise and understandable summary.”
Look up the research on that topic. Ask any clarifying questions, such as whether or not I am interested in a specific grade band or age-range of students.
Next, you will generate a concise summary of key findings from the most relevant empirical research studies and academic articles related to the topic.
The summaries should use language that is easily understandable and include specific actions that educators can actually use in their work with students. Make every word count. The summaries should be highly dense and concise yet self-contained (i.e., easily understood without access to the research articles).
Cite your sources that you used to generate the response in MLA format. Don't return a response if you do not have access to enough information. If you don't have enough information to return a reliable response, let me know this.
Note how the prompt both gives the AI a persona as well as primes it to ask users a series of questions to refine the input. You might also specify certain sources (e.g., journals, research clearinghouses, etc.) that you want to limit the AI’s landscape review to.
Keep reading with a 7-day free trial
Subscribe to Empathy in Bytes to keep reading this post and get 7 days of free access to the full post archives.