I wrote a paragraph on that in the post. I predicted a publication bias in favor of positive results, assuming the community is not biased on the particular issue of vaccines & autism. This prediction is probably wrong, but that hypothesis (lack of bias) is what I was testing.
I said “surprisingly low” because of publication & error bias.
Which way do you think publication bias on the issue goes, anyway?
I wrote a paragraph on that in the post. I predicted a publication bias in favor of positive results, assuming the community is not biased on the particular issue of vaccines & autism. This prediction is probably wrong, but that hypothesis (lack of bias) is what I was testing.