Think of interviews as a test…they have something to measure, questions to ask, and answers to score.

Something to measure: the more data the better the interview coverage (both depth and breadth)…data can come from job descriptions (so-so), managers (so-so) or thorough job analyses (most complete)

Questions: the better the question technique the more accurate the answer…techniques can be easy to fake (open-ended and closed-ended) or difficult (behavioral or situational example)

Answers: the more related to the job, the more accurate the answer…answers range from ‘nobody didn’t say nothing wrong’ (blecch!), sounds-like something I would expect (so-so), relates to something similar on the job (better), or sounds like exactly like what the job would requre (complete job-specific behavioral example).

Still, until we can find a way to read minds, interview effectiveness will be suject to lying and personal interpretation.

Why aren’t good interviews followed? In my experience:
1) recruiters have little or no incentive to screen-out more candidates
2) it’s hard
3) someone else lives with the consequences of a bad hire
4) managers are resistent to play

I, too, am amazed how few companies really put any thought into this process.
Jon

1) learn explicitly what competencies are required for the job and define the business necessity for them
2) communicate clearly what the job will be like
3) use only tests, interview questions, application methods, simulations, and so forth, that are job-related and are proven to predict job performance
4) keep all sources open to ensure qualified people of every class and group have an opportunity to apply
5) use the same rigorous criteria to select managers
6) don’t allow conflicting organizational practices to frustrate employees from doing their jobs
7) provide essential training to supplement skills when necessary

Now…compare this with what usually happens:

1) recruiters work from a job description prepared by the HR department and edited by a manager who often does not do the job…
they generally ignore the real job experts (e.g., people who actually do the job)
2) people sugar-coat the job so it will sound attractive to the candidate and hide negative information
3) reject best-practices because they are too ‘hard’ or time consuming, and rely instead on unsubstantiated personal feelings and opinions formed during short conversations.
4) selectively choose sources that minimize diversity
5) promote managers based on their performance as individual contributors and, once promoted, protect even the most dysfunctional
6) hire on one set of criteria, train on another, manage on a third, reward on a fourth, and allow organizational blocks to frustrate employees
7) try to train the untrainable or change core human behavior

Ready for the challenge?

Or a step further, if it could be measured (on employee evaluations or yearly reviews) how well they performed, giving extra credit for longevity perhaps.

Anyway, to hear the ‘Psycho-babble’ of ‘If you were a tree, what color would you smell like’ was certainly worth it for now.

Not only is anything you do to narrow down your pool considered a test for validity purposes, it’s considered a test LEGALLY (e.g., per the Uniform Guidelines, per Connecticut v. Teal).

In addition to matching your test with your measure of performance, the other big MUST DO is matching the test to the type of job. For example, if you need someone with good writing skills, why on earth are you just using an interview??