In a TV-documentary I once saw there was a prosecutor who basically said (smiling): “Yeah, I know that he fitted the criteria for insanity, everyone knew, it was so obvious. I didn’t think we’d succeed with our strategy but we had to build this *narrative* that he was sane.” And she seemed to be totally fine with it, she did her job and won, so that’s something to celebrate, isn’t it? Well, I don’t know. While she succeeded in her job, shouldn’t trials be about evidence and justice instead?
In science, I think this is
even more complicated because there isn’t necessarily a representative for
every side of an argument. While there of course is peer-review (which serves
that purpose), peer-reviewers (most of the time) can’t know what is not
reported, what is not part of the narrative that the authors tell. They don’t
go and dig failed preliminary studies or replications out of the authors’
file-drawers; they don’t search for evidence against the authors claims.
Therefore I think that in
research it is even more important that the full story, not just the prettiest
element of it is told. However, that is not what we are taught. In 2015 I
attended a Summer School for PhD-students. It was very interesting and I
learned a lot, however we also were told that we have to tell a story, report
the aspects that are most convincing and interesting to tell. While I
understand and agree with some parts of that (data can be shown in confusing or
less confusing ways) I think that this advice is mostly understood (and
probably meant) as “out of the many things you could tell about your study/your
data, you’ve to pick the most interesting part and choose the aspects of the
data that support it, while ignoring the rest”. I think that this is wrong.