There is a great example in the book “Freakonomics” by Stephen Dubner and Steven Levitt. They look at massively large databases and extract information using statistical methods.
My favorite is the example of dating and what people look for in a partner. Dubner and Levitt looked at dating sites (I forget which. Match.com, Cupid. etc.) and collected data on ~75,000,000 matches (it might be 90 million) and whether they were “successful” or not. Did the participants rate the date successful? Did they go back to the fish pond for another try? Did they go on a second date?
In their profile, women said things like: the most important characteristics in a man were things like personality, sense of humor, likes walks in the park, etc. Men said similar things. But when the data from 75 million dates was reviewed and correlated, the results were surprising. For women, which factor had by far the strongest correlation for a successful match? The man’s income! Yep! Income. It was more important than weight, height, age, anything! “Walks in the park” were not even on the list.
For men, which factor had the highest correlation to a successful date? The woman’s body weight. Next was income. They even figured out how much 10 pounds was worth in the woman’s income. (I think it was ~$50,0000. Meaning the heavier woman had to earn $50K a year more than her thinner counterpart to get the same success rate. Hey, at least guys weren’t totally one dimensional!)
You might not like the answers but with ~75 million matches they make a good psychological case. They present the data, list the sources, explain the analysis.
You can even repeat the tests yourself in class.
I guarantee yours will be the most interesting paper of the bunch.