I've been doing research to try to give a cogent reply with proper citations about why that study is bullshit, but it's difficult. The world of social psychology research is complex, and they have a nasty habit of cloaking made up bullshit with impressive sounding terms. For example, a "power analysis", which is a major part in determining how large a sample size should be, is essentially gut feelings and guesswork. The general idea is that if your methodology/test has high "power", then it will provide data correctly indicating to reject the null hypothesis more often, which therefore means you can use fewer people (lower sample size) to conduct your study. This is desirable because it makes research cheaper and faster. However, if you fuck up the power analysis, you may use too few people and your data is effectively irrelevant. There is no such thing as a sample size which is too large, except that it could identify the existence of covariables that you hadn't anticipated (e.g. perceived sexism could depend on gender and location and age and field, and any number of things not thought of or listed here), which is good for science, but bad for researchers. No researcher wants to do a study involving 10,000 participants only to publish "welp, we couldn't reject the null hypothesis."
In this particular paper that archi is clinging to as gospel (even though it's unique research and there are no replications yet, or attempts to replicate that I'm aware of), they mention having done a power analysis and therefore their sample size is good enough. They don't go into the details of what their power analysis was, or why it was sufficiently thorough. We effectively have to take them at their word. That doesn't fly with me, and shouldn't for most people. I've barely scratched the surface in learning what goes on in professional research, and I can already see this paper is missing a lot of details it should have provided.
Maybe sometime in the next day or two I'll have brushed up on my statistics from nearly a decade ago enough to be able to dissect their analysis of the small amount of data provided. Ideally I would find their raw data and do the analysis myself to ensure I got the same answers, but I don't know if I'll even be able to find that. Kind of wish Sath would weigh in on this shit, as he has first hand experience dealing with research.
XI Wiki



