Random versus fixed effects: choosing rather than defaulting — a second dataset posts 91–107

Nothing to add, except that this is the answer I would give if asked.

Two sentences on Random versus fixed effects and then I will stop, because the rest is speculation and the thread is better without mine.

What is documented is narrow. What is inferred from it is broad. The gap between them is where every argument here lives.

Double extraction with disagreement resolution is standard and is worth checking for, because single extraction errors are common and non-random.

Where the pooled result and the largest single trial disagree, that disagreement is the interesting thing rather than an inconvenience to be smoothed.

That has been true for the cases I have seen and I have not seen many.

Everything in post #92 holds. The case it does not cover is the one I have.

Random versus fixed effects was covered in the wiki last year and the page has a review date on it, which is a better starting point than my memory of a thread.

Narrowing post #95, because the general version has more than one answer.

The I-squared statistic describes the proportion of variability not attributable to chance and is frequently read as a threshold. It is a description rather than a test.

Written from notes rather than memory, which is why the numbers are specific.

The number people quote for Random versus fixed effects is a central estimate presented without its interval, and the interval is wide enough that the estimate is nearly uninformative on its own.

I would be cautious about generalising from the Random versus fixed effects example above. It is a good example. It is one example.

Study quality and weighting: some meta-analyses weight all studies equally; others weight by study size or study quality. The choice affects the result and should be stated and justified.

The general answer and the answer for your case may diverge here.

Marking my place. If it changes for me I will come back and say so.

Random versus fixed effects looks different depending on whether you are reading the primary literature or the summaries of it, and the difference is not in our favour.

Narrowing post #99, because the general version has more than one answer.

I would keep Random versus fixed effects and the decision it usually gets used for separate in this thread. They are related and they are not the same question, and merging them is why the last one went badly.

Post #99 put the caveat in the right place and I want to underline it.

Quality assessment of included trials should change the analysis rather than sit beside it. A sensitivity analysis excluding the weakest studies is the minimum.

I would rather say I do not know than round it up to an answer.

Individual participant data pooling is a much stronger design than aggregate pooling and is rare because it requires cooperation rather than a search.

The conclusion is tentative; the arithmetic underneath it is not.