selectivity
The statistical problem that observed groups are non-random samples (selected on characteristics or circumstances), biasing comparisons of their outcomes.
As defined in Claudia Goldin
Where it comes up · 1
- Claudia Goldin2 claims
Most important claims · 2 of 2Search all claims for “selectivity” →
- The Black male wage gap is a highly selected statistic because measured outcomes differ dramatically depending on whether one includes the entire population, only the non-incarcerated, or only the never-incarcerated, and these selection effects reflect some of the nation's largest social ills.Claudia Goldin
- The decision of whether women work at all (the extensive margin) is the larger margin and is more responsive to incentives, and because women who are in particular workplaces are selected on personal characteristics or their spouses' characteristics, selectivity complicates interpretation of observed earnings patterns.Claudia Goldin