Evaluating Information
No matter where you get your information, you need to make sure you critically evaluate each source to ensure it’s appropriate for your research! Many publications have a particular bias or agenda, which may not be obvious at first glance.
Here are a few criteria that could help you in your evaluation:
Authority
- What are the author's credentials and affiliation?
- Who publishes the information?
Accuracy
- Based on what you already know about the topic or from reading other sources, does the information seem credible?
- Does the author cite other sources in a reference list or bibliography, to support the information presented?
Scope
- Is the source at an appropriate comprehension or research level?
There are other criteria to consider as well, such as currency, objectivity, and purpose. For more information, see UBC Library’s Evaluating Information Sources.
Citing Sources
- How to Cite Sources - UBC LibraryQuick guides to common styles such as APA, MLA, and more.
- Learning Commons - Irving K. Barber Learning CentreHelp with technology, citing, study tips, and more for UBC undergraduates.
Citing AI Generated Text
Before using ChatGPT for assignments, refer to UBC's Academic Integrity page and be sure to check with your instructor to find out if AI tools are permitted.
Is ChatGPT trustworthy?
Responses can reflect biases from the people who wrote the original text used in the training dataset. ChatGPT was trained using a massive dataset of text written by people and pulled from the Internet. ChatGPT is not connected to the Internet and the data used to train it was collected prior to 2021. Information may contain errors and be outdated.
Fake citations
It has been well noted that ChatGPT and similar AI tools generate false citations. ChatGPT makes stuff up!
Privacy
OpenAI (the company that designed ChatGPT) collects data from ChatGPT users. Their privacy policy states that this data can be shared with third party vendors, law enforcement, affiliates, and other users. Input into ChatGPT cannot be deleted. Why does this matter? If you ask ChatGPT about sensitive or controversial topics, that data cannot be removed or deleted and others may have access to it.
Some content on this page has been adapted or adopted from the open licensed An Instructor's Guide to Teaching & Learning with Technology @UNBC by UNBC CTLT
Evaluating for AI Generated Content
There are some detection tools that can help determine whether or not text has been generated by a human, AI or both. Here are some examples:
All sources you find on the internet should be evaluated carefully whether they are AI generated or not. Using the SIFT method is a helpful way to determine if the source you have found is reliable and expert information.
"SIFT (The Four Moves)" by Mike Caulfield, re-used and adapted under CC BY 4.
There are four moves to help you evaluate information you find on a website. Using the Library search, Summon, helps find research articles, books and more to help verify information.
- Stop. Do you know the website or where the information is coming from? Can you verify the claims in the website? Before you click "share" make sure you have verified the information.
- Investigate the Source. Who is providing the information? What is their level of expertise? Do they have an agenda that might influence what is said and how the information is presented? Using the Library search, Summon, can help verify the information and provide multiple reliable sources.
- Find Better Coverage. Look into multiple sources that can help verify the claim. What are the experts saying about the topic? Where else has the story been covered?
- Trace claims, quotes, and media back to the original context. Trace the claim you find on social media, or a news clip, or a website you land on back to the source to get the context for the claim.