Date: 2023-02-20
Have you ever noticed how people frequently report struggling with statistics, but don’t usually do the same regarding other (arguably) equally difficult subjects? I describe what I’ve seen, and provide some possible explanations for this phenomena. I’m really interested in knowing what are your thoughts on this, so please open an issue in this blog’s GitLab, or reach me in LinkedIn with your ideas.
Everyone claims to struggle with statistics and to be ignorant about them. To be clear, statistics, can be a complex field, and navigating the lively debates in the field (e.g., frequentism vs bayesianism) is hard. Still, from my experience people complain about struggling with statistics much more than they complain about understanding their research topics. To my naked eye their fields of study don’t seem that much easier, or simpler, than statistics. Maybe this is just be being a victim of the curse of knowledge, as I like statistics, but I think it is more than that. It’s also worth noting that this seems like an instance of the below average effect, the opposite of the, better-known, better than average effect. Still, that is not an explanation for why this occurs, just a suggestion for how to label the effect.
I ranted about why I reckon most people are struggling with something else, when they claim to be having a hard time with statistics, in a previous post. Still, even if you agree with my points, that just means people blame statistics for other problems they’re having, it doesn’t account for why they blame statistics, and not some other innocent topic.
Let us try to come up with some possible explanations for this phenomena then. By us, I do mean us, I’m going to put forward some hypotheses but I would like to see what you think. Feel free to open an issue in this blog’s GitLab, or reach me in LinkedIn.
This uninteresting hypothesis, posits that people started to report struggling with statistics, and that made others comfortable sharing the same. Since the same didn’t happen for other areas of expertise, statistics remain the only area about which people report doing that. This hypothesis as little explanatory power given it doesn’t account for why people started reporting struggling with statistics, specifically, but not other areas.
This isn’t really an explanation as to why people only report struggling with statistics. Instead, it’s merely an hypothesis about why the effect persists. This hypothesis states that the effect persists because it is hard for people the people who actually like statistics, and feel comfortable working with them, to come out and say that.
According to this hypothesis, people actually struggle with statistics because university curricula provide insufficient training in this area (but not on others). I do think this is true, in most cases, but I also think there is more to it than that.
Statistics can be considered a field of math (though I recall my high-school math teacher saying that was up for debate). Math is already victim to many prejudices, with a reputation for being harder than the other classes. Maybe people are simply perceiving statistics as math, and being equally prejudiced towards it.
Unlike many other areas of expertise, statistics, in what concerns psychological sciences and related fields, is a practical skill. Not only must we understand the underlying theory, as we must know how to operationalize and test our hypothesis statistically. Doing data analysis, and having a hard time with it (e.g., not being able to perform an analysis in a given software) may provide people with more feedback, than when people are writing a literature review. Statistical software gives out errors, but word processing software won’t complain if miss important points, or misinterprets some key findings, in your literature review.
If we think about it the above explanations are not really mutually exclusive, so all of the above can be true (or partially true), or some combination of them at least.
I’m interested to know what you think, please open an issue in this blog’s GitLab, or reach me in LinkedIn, with your ideas.
Thank you so much for reading!