An exercise that teaches ChatGPT’s limitations. This exercise empowers students to verify the information AI generates, fostering responsible AI use.
Ask ChatGPT to generate a list of 4 academic sources on a topic of your choice, and then evaluate the credibility and usability of those sources.
What is the topic you chose?
What 4 citations were generated? (Paste the citations here)
THEN complete the following:
1. Are the citations actually real? Does such a journal/website/book exist? State which are not real and which are real. State whether any website used in a real citation where you found it is credible and why.
2. State where those specific real citations are available full-text (check our library databses too). List the names of the places you found them (for example, name of such-and-such webite, name of database , etc...).
3. Check the credentials of the lead author by doing a google search of their name in quotes. Are they trained in the field of the topic? State their credentials and/or academic degrees.
4. Now run their name (in quotemarks) in a library database (like ProQuest or Ebscohost), use a drop down to search for AUTHOR - do they appear? IF YES, What are their other article/s (provide the permalink URLs) about?
5. Now run a search for your same chosen topic in a library database. What are the top four most relevant (provide the four permalink URLs)? Note if they match any of the original four generated.
Bonus 1 point: Talk about paid and unpaid access to this AI tool (look at pricing for different versions on the Chatgpt website) and how YOU think it might affect what you find in any tier of paid/unpaid access.
This assignment tracks to the ACRL Information Literacy framework:
"Information has Value"