Researcher bias/allegiance are synonyms.
Basically the simple idea that investigators/researchers tend to hold views that constitute allegiance to particular ways of seeing the world that these, through both conscious decisions and through less deliberate processes, will affect how they design, conduct and report their work and this may be to the extent that the discussion and conclusions are biased.
Details #
Almost certainly applies particularly strongly to research into psychological and psychosocial therapies with a strong tendency for it to bias findings toward the therapy the researcher/investigator favours. There was a very strong association between the allegiances of primary researchers and their findings in early therapy controlled trials when research teams were usually unified in their allegiances. That led to advice that teams running evaluations should have a mix of therapy allegiances and probably this does help minimise the effect. One interesting suggestion is of “triple/treble blind” trials or “blinded analysis” where in the traditional comparative trial the results are encrypted such that the researchers and, above all the data analysts and authors of the report are only unblinded to which arm of the trial was which when the report is signed off and submitted.
That still leaves room for bias to creep back in through suggestions from peer reviewers and from the authors in their amendments to the report following peer review. However, I have yet to see triple blinding for a therapy trial. Odd when you think about the logic of the randomised controlled trial and particularly of the double blind randomised controlled trial! Not only have I not seen a triple blind trial in our field, I’ve never seen a suggestion of quadruple blinding (my invention) in which the blinding of the arms of the trial would continue in place until the peer review process was completed.
Try also #
Blinding
Controlled trials
Double blind randomised controlled trials (DBRCTs)
Randomised controlled trials (BRCTs)
Research(er) equipoise
Methodology
Chapters #
Not specifically mentioned anywhere but chapter 10 is about how we can try to minimise these effects
Online resources #
None
Dates #
First created 6.xii.23
