How to Evaluate Sources
Not all information is created equal. Some sources are more trustworthy and more appropriate to academic research than others. Learning to evaluate information sources is a key skill both in research and in life.
While there are many criteria that you can use to judge a source, we recommend looking at Author, Review, Date, Bias, and Sources. Below you’ll find more information on how to evaluate using those criteria.
Author
When you’re looking at a source, pay attention to who wrote it. Are they qualified to write about this issue? Should you trust them? Check the authors of books and articles and especially those of blogs and websites. Here are some factors to consider:
- Is the author who they say they are? See if you can verify their credentials.
- For websites, what kind is it (.org, .edu, .gov, etc.)? Is more information about the group available?
- Is an organization or group listed as the author? If so, what are their credentials?
Review
Books, articles, and newspapers all undergo some form of review process. Scholarly books and articles will be reviewed by editors and experts in the field to check for accuracy and to assess the research methodology. Newspapers are reviewed by an editor. Many sources you can find online have not had any review. That means that there has been no one to check for accuracy. Here are some things to consider when looking at a source:
- Has the information been reviewed? Was it published in a journal or by a reputable publisher?
- Can you get more information about the publisher? For example, for a journal, can you check their website to see who runs the journal? Are they affiliated with a college/university or professional association?
Date
The currency of information is essential for some types of research and less so for others. Historical information that reflects people and events of the past is relevant to research in many fields. In other fields such as health care, legislation, and finance, current information is most important. Here are some things to consider:
- When was the information published or last updated?
- Have newer articles been published on your topic?
- Are links or references to other sources up to date?
- Is your topic in an area that changes rapidly?
Sources
Scholarly sources will always list the sources used, generally in the form of a bibliography. Other information types, like websites and blogs, might list sources but may not. Be skeptical of information that doesn't list a source. When you review sources, check:
- Does the author cite the work of others?
- What kinds of sources are listed?
- For online sites, are the links current or are they dead ends?
Bias
All information can have bias. Being aware of what a source is trying to convince you of and why can help you decide whether it is trustworthy or useful. To evaluate the bias of a source, consider:
- What is the intent of the source? Does it want to persuade you? Does it want you to buy something?
- Are there ads? How to they relate to the content?
- Is the author presenting fact or opinion?
- Is the language used impartial?
Context
The criteria discussed so far are used to evaluate a single paper, webpage, book, or other source. But since sources do not exist in a vacuum, we must also consider their context: how do they relate to other works on the same topic? Do they reach a similar conclussion to other sources or are they outliers?
If you find that one (or several) of your sources draws unusual conclussions, that doesn’t mean that it isn’t a good source—in fact, it may be an especially important one.
One possibility is that your outliers aren’t really outliers—you may have just happened to come across more papers on one view than another. In this case, a little more searching (possibly using a citation search to find articles that have cited the ones you have in hand) may turn up more sources like your outliers.
If the sources really are outliers, ask yourself why. Do they represent a distinct viewpoint on the topic (e.g., practitioner vs. theoretician)? A different methodology (e.g., statistical vs. qualitative)?
Do the authors acknowledge alternate viewpoints or conflicing results? Do they cite other major papers on the topic or does their work seem to exist in a vacuum? A good “outlier” author will acknowledge the work of more mainstream researchers, even if they disagree with them.
AI Chatbots
LLM-based chatbots are a recent and rapidly evolving technology, and it can be tough to figure out how (or if) they can help you with your research. It’s important to remember that services like ChatGPT and Gemini aren’t actually intelligent. They are trained on large amounts of text and programmed to give human-like answers, but they don’t actually understand or generate new information. AI chatbots can’t produce new research and can even struggle to accurately summarize existing information. Research shows that they often “hallucinate” or introduce false information. It’s important to be very cautious and use your critical thinking skills when dealing with AI.
If you really want to experiment with using AI in research, here are a couple of ways it could be helpful:
- Use AI to generate a list of search terms and ideas
- Ask AI to organize and find themes in your research notes
- Use AI tools to find related citations