So let us assume in whatever field there is a difference between 'life' and 'research'.

Consider the following bias:

- in a research project, cases are added to one group which would in 'real life scenarios' not be identified as such.

For example a patient with a rare disease is added to a cohort for research purposes, while in real life the diagnosis is not strong enough to justify dangerous therapy.

The research cohort is inflated (possibly to allow 'stronger' statistics or reach a minimum group size), yet the over-included cases would better be suited as control.

Obviously this problem is rather simple. I am asking:

- is there a name already for that kind of bias?

- can you name a citation or researcher?

Thanks a bunch,

Stefan