Applications statistics

An employer counting applications received per opening, a researcher counting responses to submitted applications, and a job seeker counting their own applications are measuring three different quantities. All three get reported as an application-to-interview rate.

Denominators do most of the work here. A rate is only meaningful with the population it was drawn from attached, and most circulating figures have been separated from theirs.

This is also the area where averaging is most tempting and most wrong. Combining an employer-side rate with an audit-study contact rate produces a number that describes nothing.