mainly data collection, cataloging and cleaning

Maybe some logistic regression if we can find people who do and
*don't* apply for jobs.  Or people who do or *don't* stay in their
positions.

Hot deck imputation?

What kind of people who apply for jobs are most like those who don't?

  - those who apply latest?

What kind of people who stay in jobs are most loke those who don't?

  - much less clear