Selection bias is a distortion that arises when the matters included in an analysis are not representative of the population the conclusion is meant to cover. It can come from how data were collected or from who chooses to bring a case.
Examples include analyzing only matters that reached judgment, which excludes those that settled early, and analyzing only matters a funder accepted, which excludes the ones it declined. Conclusions then apply to the filtered group and not to the whole.
The remedy is to define the reference population explicitly and to state what was excluded and why. Where a bias cannot be removed, the claim is scoped to what the data support.
Selection bias is not removed by a larger sample. More data from the same filtered source repeats the same distortion.
Working a file that turns on this term?
Working through a diligence process?
Institutional partners evaluating a position against this platform's outcome and duration models are welcome to reach out directly.