Ask HN: How do you find quality research on unfamiliar topics?

[xalava]

I was asked a great question today: "How do you find good studies on a topic? Is the number of citations an indicator of quality?"

Over the years, I've learned many tricks, but there isn't a straightforward answer. This is especially true for someone who isn't working full-time in a field but is interested in finding solid information and arguments.

What would you recommend?

[dredmorbius]

Start with a good canonical reference: a textbook, a college syllabus (though these are getting harder to find), an encyclopedia article. Look at the references.

Even if you're looking for recent material, those references should be useful as principle studies or background will itself be referenced in more recent research.

Citation counts themselves are at best a highly uneven quality indicator. The real test of research is replication: a study turns up phenomena which are independently validated by other researchers. This of course takes time.

Good current research tends to come from teams and institutions which have produced substantive work in the past. So the references / citations approach I'd opened with should also give you names of individual researchers and/or institutions which have had high-impact work in the field Their more recent work will likely also be of interest. This approach will tend to underrepresent novel work from new teams ... but that work is also often less likely to be significant or robust.

If you're a practitioner in the field yourself, and have the equipment / means to test findings, validating results yourself is another option. Mind the various biases which can come into play here.

[begemotz]

this is probably field dependent, but I would say:

- don't focus on individual studies but try to find meta-analyses.

- if not available, then review articles.

- if not available, then individual papers.

# of citations is probably a closer proxy for influence rather than "good" (if you mean quality). Of course, one hopes that there is a strong correlation between the two.

More in the weeds would then assess the impact factor of the journal that an article is published in as well as other 'altmetrics'.