In November of 2013 Coca Cola hosted a working group in the UK to discuss best practices around the utilization of assessments in the employee recruitment and development processes. Several heavy hitters attended, including Archer Daniels Midland (where yours truly is employed), Caterpillar, Citi, IBM, and Microsoft. This was a good discussion, and I recommend checking out the full working group summary here.
With that said, I’m curious to get reader’s thoughts on one potentially curious assessment practice that a few of the companies present utilized:
Several partner companies indicated that, depending upon the risk associated with a given role, an individual could be disqualified for a position pending the outcome of the assessment process. While the lack of specific technical skills was often the disqualifier, the lack of specific behavioral competencies was also used to disqualify candidates.
Assessments as defined in this working group are tests, interviews, simulations, assessment centers, and background reviews. And some of these tools – like interviews and background reviews – have well documented predictive value when testing for certain attributes under the right conditions. What would be odd to me here, however, would be DQing a candidate on the basis of a behavioral assessment that is a simulation or test – both of which are artificial and self-report environments, respectively. And to be fair to the companies attending this working group, while some of the assessments they mention using are self-report (i.e. Harrison and DISC), the summary report doesn’t explicitly say what types of assessments are being used in cases where candidates are being disqualified based strictly on the results.
But if candidates are being disqualified at some organizations strictly on the basis of a behavioral test, simulation, or exercise is that okay?
To be sure, behavioral assessments can be fantastic pre-screens for online applicants (I think they’re as good a tool as any to thin a large applicant pool). I’m also all for using technical assessments to disqualify candidates that don’t have the skill set to do a job…
…But should a score on a behavioral assessment ever be enough to end an internal applicant’s candidacy for a role? I won’t dismiss this idea out of hand, but I’d think such an assessment would need to have extremely high predictive value to justify such a decision. To this point, even the best self-report qualitative tests are vulnerable to some degree of incidental subjectivity – and 1. I once had a manager recount to me a story where his boss chastised him for getting the “wrong” personality type on an internal test. The manager then re-took the assessment, this time gaming the system to get the personality type his boss wanted him to score. Afterwards, he said to his boss “Happy now?”.perhaps even respondent faking of answers. 1 Furthermore, a test taker may exhibit a recency effect when answering a self-report questionnaire or participating in some sort of simulation (i.e. if they’re having a bad day their answers might reflect that).
…So obviously I’m not a big fan of placing definitive (i.e. advance or DQ) weight on most behavioral assessments. Frankly, I think there are just too many confounding variables in play for such a decision to be valid.
Perhaps I have this wrong, though? As much as I enjoy analytics, in my roles to-date I’ve admittedly had little experience administering assessments or analyzing the results afterwards. So there could be a lot of conditions I’m making a big deal about which are actually perfectly controllable under the right test taking conditions and with the right questionnaire wording.
Do you have experience administering behavioral assessments internally? If so, please let me know what I’m missing here in the comments section below.