[IPAC-List] Seeking Taylor-Russell formulas

[mhammer at 295.ca]

Utility analysis is relevant for that period where those who *used* to be
candidates are now incumbents.  That is not to dismiss its relevance or
importance, but it tends to ignore retention as a critical aspect of
evaluating test-utility.  The predictive validity and utility of tests are
for identifying candidates who will accept your offer AND stick around,
such that the utility and added value can be realized.

So, without wishing to derail or distract from the thread, I was curious
as to whether there is research or reports that attempt to integrate
validity/utility with prediction of retention.  I ask this with the
understanding that probability of non-retirement voluntary departure often
decreases with tenure, just as evidence of utility *increases* with
tenure.

Grosso modo, the ideal is for any pre-employment testing to identify not
only those who will add value for the employer, but do so for a period
long enough to justify that investment in assessment and selection.

Feel free to ignore.

Mark Hammer
Ottawa
(now retired)