Project Design and Paper Review

My Week In Three Words: Routine, Design, and lots of Reading

What I’ve been working on: This week I’ve been continuing paper review, drafting a survey design, and drafting a document with our paper outline. I’ve been re-reading the related works sections of papers to find other work donw around content moderation and interestingly enough found a paper with a very similar study format to what we are planning to do. The title is Each to His Own: How Different Users Call for Different Interaction Methods in Recommender Systems. Even though the study was on recommendation systems, the format of having this interface to test, getting users to play around with it, and then asking survey questions is very similar to what we’re hoping to do. At our team meeting, we also discussed whether we want this to be a between-subject or within-subject study as well as improving on our conditions that we’ll test. There are a few conditions I’d like to discuss more with the group. “Threshold” vs “Probability” is a confusing concept to me. It’s supposed to be that threshold is “the level of toxicity of a comment” whereas probability is “the likelihood of that comment being toxic.” But I’m wondering if there’s a benefit to just combining them? I’ve also been going through from LaTeX tutorials to get familiar with it for when we’ll start writing the paper.

What I’ve learned:

1. Make your resume on LaTeX–the formatting is soo much easier!
2. If you’re looking to read up on similar work/to get ideas at the beginning of a research project, looking at papers cited in the related works section is a great way to start. An additional plus is that you’ll go in having more context than if you were to do a keyword google scholar search.
3. The google scholar chrome extension (I think it’s called the Google Scholar Button) is pretty great ;)

What I’m curious about:

1. What are the benefits of doing a mixed study (of both between and within subjects)?
2. How do you decide between a between subjects and a within subjects study?