How to make ChatGPT provide sources and citations

We Keep you Connected

How to make ChatGPT provide sources and citations

 

Jonathan Raa/NurPhoto via Getty Images

One of the biggest complaints about ChatGPT is that it provides information, but the accuracy of that information is uncertain. Those complaints exist because ChatGPT doesn’t provide sources, footnotes, or links to where it derived information used in its answers.

While that is true for the GPT-3.5 model, the GPT-4 and GPT-4o models provide more citation resources. While GPT-4 is only for paid subscribers, GPT-4o is available to both free and paid subscribers, although free users get fewer citations and less detail than users with a ChatGPT Plus subscription.

Here’s how ChatGPT describes the approach: “GPT-4o in free mode provides basic and essential citations, focusing on quick and concise references to ensure information is traceable. In contrast, GPT-4o in paid mode offers enhanced, detailed, and frequent citations, including multiple sources and contextual annotations to provide comprehensive verification and understanding of the information. This ensures a robust and reliable experience, especially beneficial for users requiring in-depth information and thorough source verification.”

Even with the provided citations in GPT-4o, there are ways to improve the approach. Let’s take a look.

How to make ChatGPT provide sources and citations

If you know how to prompt ChatGPT, it will give you sources. Here’s how.

 

To start, you need to ask ChatGPT something that needs sources or citations. I’ve found it’s better to ask a question with a longer answer, so there’s more “meat” for ChatGPT to chew on.

Keep in mind that ChatGPT can’t provide any information after January 2022 for GPT-3.5, April 2023 for GPT-4, and October 2023 for GPT-4o, and requests for information pre-internet (say, for a paper on Ronald Reagan’s presidency) will have far fewer available sources.

Here’s an example of a prompt I wrote on a topic that I worked on a lot when I was in grad school:

Describe the learning theories of cognitivism, behaviorism, and constructivism

 

 

 

 

This is where a bit of prompt engineering comes in. A good starting point is with this query:

Please provide sources for the previous answer

I’ve found that this prompt often provides offline sources, books, papers, etc. The problem with offline sources is you can’t check their veracity. But it’s a starting point. A better query is this:

Please provide URL sources

This prompt specifically tells ChatGPT that you want clickable links to sources. You can also tweak this prompt by asking for a specific quantity of sources, although your mileage might vary in terms of how many you get back:

Please provide 10 URL sources

In our next step, we’ll see what we can do with these sources.

 

 

 

ChatGPT and most large language model AIs respond well to detail and specificity. So if you’re asking for sources, you can push for higher-quality sources. You’ll need to specify that you need reliable and accurate sources. While this approach won’t necessarily work, it may remind the AI to give you more useful responses. For example:

Please provide me with reputable sources to support my argument on… (whatever the topic is you’re looking at)

You can also tell ChatGPT the kinds of sources you want. If you’re looking for scholarly articles, peer-reviewed journals, books, or authoritative websites, mention these preferences explicitly. For example:

Please recommend peer-reviewed journals that discuss… (and here, repeat what you discussed earlier in your conversation)

When dealing with abstract concepts or theories, request that ChatGPT provide a conceptual framework and real-world examples. Here’s an example:

Can you describe the principles of Vygotsky’s Social Development Theory and provide real-world examples where these principles were applied, including sources for these examples?

This approach gives you a theoretical explanation and practical instances to trace the original sources or case studies.

Another idea is to use sources that don’t have link rot (that is, they’re no longer online at the URL that ChatGPT might know). Be careful with this idea, though, because ChatGPT doesn’t know about things after January 2022 for GPT-3.5, April 2023 for GPT-4, and October 2023 for GPT-4o. So, while you might be tempted to use a prompt like this:

Please provide me with sources published within the past five years.

Instead, consider using a prompt like this:

Please provide sources published from 2019 through April 2023.

And, as always, don’t assume that whatever output ChatGPT gives you is accurate. It’s still quite possible the AI will completely fabricate answers, even to the point of making up the names of what seem like academic journals. It’s a fibber. It’s a sometimes helpful tool, but also a fibber.

 

 

 

Keep this golden rule in mind about ChatGPT-provided sources: ChatGPT is more often wrong than right.

Across the many times I’ve asked ChatGPT for URL sources, roughly half are just plain bad links. Another 25% or more of the links go to topics completely or somewhat unrelated to the one you’re trying to source. GPT-4 and GPT-4o are slightly more reliable, but not much.

For example, I asked for sources on a backgrounder for the phrase “trust but verify”, generally popularized by US President Ronald Reagan. I got a lot of sources back, but most didn’t exist. I got some back that correctly took me to active pages on the Reagan Presidential Library site, but the page topic had nothing to do with the phrase.

I had better luck with my learning theory question from step 1. There, I got back offline texts from people I knew from my studies who had worked on those theories. I also got back URLs. But once again, only about two in 10 worked or were accurate.

But don’t despair. The idea isn’t to expect ChatGPT to provide sources that you can immediately use. If you instead think of ChatGPT as a research assistant, it will give you some great starting places. Use the names of the articles (which may be completely fake or just not accessible) and drop them into Google. That process will give you some interesting search queries that probably lead to interesting material that can legitimately go into your research.

Also, keep in mind that you’re not limited to using ChatGPT. Just because ChatGPT exists doesn’t mean you should forget all the tools available to researchers and students. Do your own web searches. Check with primary sources and subject matter experts if they’re available. If you’re in school, you can even ask your friendly neighborhood librarian for help.

Don’t forget that there are many excellent traditional sources. For example, Google Scholar and JSTOR provide access to a wide range of academically acceptable resources you can cite with reasonable confidence.

One final point: if you merely cut and paste ChatGPT sources into your research, you’re likely to get stung. Use the AI for clues, not as a way to avoid the real work of research.

 

 

FAQ

How do you put sources in APA format?

APA style is a citation style that’s often required in academic programs. APA stands for American Psychological Association. I’ve often thought they invented these style rules to get more customers. But, seriously, the definitive starting point for APA style is the Purdue OWL, which provides a wide range of style guidelines.

Be careful: online style formatters might not do a complete job, and you may get your work returned by your professor. It pays to do the work yourself — and use care doing it.

How can I make ChatGPT provide more reliable sources for its responses?

This is a good question. I have found that sometimes — sometimes — if you ask ChatGPT to give you more sources or re-ask for sources, it will give you new listings. If you tell ChatGPT the sources it provided were erroneous, it will sometimes give you better ones. The bot may also apologize and give excuses. Another approach is to re-ask your original question with a different focus or direction, and then ask for sources for the new answer.

Once again, my best advice is to avoid treating ChatGPT as a tool that writes for you and more as a writing assistant. Asking for sources to cut and paste a ChatGPT response is pretty much plagiarism. But using ChatGPT’s responses, and any sources you can tease out, as clues for further research and writing is a legitimate way to use this intriguing new tool.

Why are ChatGPT sources often so wrong?

For some links, it’s just link rot. Some links may have changed since many sources are at least three years old. Other sources are of indeterminate age. Since we don’t have a full listing of ChatGPT’s sources, it’s impossible to tell how valid they are or were.

But since ChatGPT was trained mostly without human supervision, we know that most of its sources weren’t vetted and could be wrong, made up, or completely non-existent.

Trust, but verify.


Latest stories for ZDNET in Artificial-Intelligence

GET THE LATEST UPDATES, OFFERS, INFORMATION & MORE