My level of flashback is "went to an incredibly conservative right-wing Catholic seminar debate-style liberal arts school" so yeah exactly this 😩

A more out-there reason, for someone like me with some visibility in tech, there's a nontrivial amount of "copying the way you described this to write a blog and claim to originate the idea" that happens in software commentary "thought leadership" 😅 different kind of scooping but really annoying

In doing applied & community-based work (e.g. I've recruited a lot of people who otherwise do not participate in software research, esp bc the samples in SRE are terribly skewed to all white male), holding some of those analyses quiet to the public I think aids more authentic answering

Yes, very easy (in my applied world where majority of the audience is gen public I don't think anyone even looks at these tho we do them but it's nice as a record) and because I have a public presence it generally helps to cut down on participants pre-knowing what connections you are proposing

OH, "validating prejudgments" is exactly what I was thinking. Like it seems to me these papers I've read are using the same analysis for 2 things 1) justifies a causal claim like 'these three things we selected are the drivers of x' and 2) says well given that 3 things->x here is what it looks like

thank you!

oh thank you this was totally one that I pulled up to read later, appreciate the validation that it's a good commentary

(this whole very essentialist "there are individual differences in gendered problem solving"/"GenderMag" cognitive styles thing is like....what is going on here? but that's a separate question I have......)

Here is a pretty thorough (one of the most thorough) descriptions I've seen in a software research paper, which includes a reference to how PLS-SEM is increasingly used in SER - arxiv.org/pdf/2409.04099dl.acm.org/doi/abs/10.1...

my friend described it as something they did (in bio) when there were totally known (like literally in the fossil record determined) limited sets of things and was also very skeptical of the examples I showed him (altho liked the method for what he did)