Political Humor and Anxiety: Journals and Dev Tools (Update #4)

Lit review work has been going well, continuing my progress in psych-related humor research. I’m looking now at articles from Humor: International Journal of Humor Research, which is an interdisciplinary journal for humor studies and generally assumes the readership doesn’t have extensive background in all of the fields involved (which is great for me). While it doesn’t focus on political humor, it does look at the more psychological aspects of humor and I would still recommend it to anyone looking into humor research.

I’m currently expanding my annotated bibliography and exploring the various methods of presenting humor and anxiety-causing stimuli to participants. Most studies use video clips, such as using movie footage of an airplane crash to cause anxiety and using video of a comedy sketch to present humor. However, there are some studies that use printed comics and joke punchlines to elicit humor, so I’m gathering a list of these studies and comparing their results to determine which method is best for my research question.

In terms of progress on the scripts for social media mining:

Twitter:

I started off using the twitteR package, since I was able to find the most tutorials for it, but when I checked the Github page, I found out that the package is being deprecated in favor of rtweet, a different R Twitter-processing package that is based off of twitteR and retain a lot of the same functions.

  • The twitteR package would truncate tweets and wouldn’t record the URLs of individual tweets. using the rtweet package circumvented these issues.
  • There are some key differences between the packages. The search_tweets function, while the syntax is similar, is missing some of the parameters used in the rtweet package, but the developer notes mentioned that was due to Twitter API changing. I really miss the date range search feature from twitteR and I’m working to replicate it in the new package by using the 30 day search function and then omitting observations outside of the desired date range.

Reddit (PRAW module)

  • Finally resolved the 403 error I kept getting! I changed a couple things, most notably my user agent string as per this forum response and I was finally able to run my script, which has gone off without a hitch.

Facebook:

  • Made a developer account on Facebook and beginning to write code using the Rfacebook package

I’m working on commenting the code so that my code will be easily understood for both myself and any lab members going forward, along with a guide for running these scripts and common troubleshooting issues. The packages that I’ve been using have other cool applications (tracking trends in social media, frequency of word use, etc.) once they’re set up, so I may create a small tutorial on how to do that as well.