Cornell University Food and Brand Lab used public records to analyze inmates' final meal choices before being put to death. I thought it was an interesting topic to research and an interesting way in which to research it. The summary of the research can be found here:
It is not surprising that vegitables and fruits were not requested often. However, I think it is interesting that almost 40% of inmates requested a branded food or drink. Comfort foods make sense for the situation. In terms of research, public record seems to be a flawless approach for such a specific study because it is fact and you are assured minimal error. There is a response or data for every person, the only problem is if the meal is entered incorrectly.