Generative AI and Research

Whether we are aware of it or not, we evaluate the information we encounter all the time. We use a number of clues to help us decide whether a piece of information is credible, including the context in which we encounter it, who the author is, and (if applicable) who shared the information with us. 

Evaluating information for a research project is a similar process but one where we may need to pay attention to slightly different criteria than when we are evaluating information in other contexts. 

Generative AI tools can help you identify what criteria to prioritize when evaluating information for a research project and help you understand how to apply those criteria. 

I am an undergraduate student writing a 3-5 page paper on the role of newspapers in the development of society for an Early American History course. Create a list of criteria I can use to evaluate the quality and credibility of the sources I find in my search for information on this topic. 

Below is the response ChatGPT gave the author of this guide in response to the above prompt. 

1. Authorship

• Who is the author? Are they a historian, journalist, or expert in Early American history?
• Credentials: Do they have academic qualifications or professional expertise in the field?
• Affiliation: Are they affiliated with a reputable university, research institution, or publication?

2. Publication Source

• Type of Source: Is it a peer-reviewed journal, book published by a university press, or a primary source (e.g., an early American newspaper)?
• Reputation: Is the journal, publisher, or database well-regarded in history or academia?

3. Date of Publication

• Is the publication date appropriate for your topic? Historical research often relies on both primary sources (from the period you’re studying) and recent scholarship to understand current interpretations.
• Avoid outdated secondary sources unless they are considered seminal works.

4. Accuracy

• Citations: Does the source include references or a bibliography to support its claims?
• Consistency: Do the facts align with what you’ve read in other reputable sources?

5. Relevance

• Does the source specifically address the role of newspapers in Early American society or regional identity?
• Is it focused on your time period (e.g., Colonial America, Revolutionary Era, Early Republic)?

6. Perspective

• Bias: Is the source written with an obvious agenda or bias? (For example, partisan newspapers from the period may reflect specific political perspectives.)
• Balance: Does the author present multiple viewpoints or only one side of the story?

7. Primary vs. Secondary Sources

• Primary Sources: Are they original materials from the period (e.g., newspaper articles, letters, or advertisements)?
• Secondary Sources: Do they provide analysis, context, or interpretation of primary sources?

8. Intended Audience

• Academic or Popular Audience: Is the source aimed at scholars, students, or the general public? Scholarly sources are usually more rigorous.
• Accessibility: Is the language academic and precise, or overly simplified?

9. Publisher or Database

• Trusted Platforms: Are you finding the source in a reputable academic database (e.g., JSTOR, Project MUSE, America: History and Life)?
• Publisher Reputation: Is it published by a university press or other respected academic publisher?

10. Relevance to Your Thesis

• Does the source provide information or perspectives that directly support your argument or help answer your research question?
• Are the examples and case studies in the source specific to newspapers' societal impact?

ChatGPT has given a lot of information here. Trying to apply all these criteria to every source you find would be overwhelming. To help with this, you can either ask ChatGPT to refine the list for you or you can apply your own thinking by considering: 

• How many of these criteria should any one source meet in order to be considered of adequate quality or credibility for your research? 
• Which criteria are the most important for judging the quality or credibility of a source for your research? 
• Which criteria are “nice to have” but not required in order for a source to be considered of adequate quality of credibility for your research? 

As you apply these criteria, you may change your mind about which are most and least important or find that you are using criteria that ChatGPT didn’t even suggest. That’s okay! Any list an AI gives you should be used as a general guide. There should always be some flexibility in the evaluation process.