Sunday, 28 July 2013

Accurate Performance Predictors?

As a member of the HRINZ LinkedIn group, I was reading a member post the other day that was really fascinating. Posted by Anna Sage of Sage Advice in Wellington, it detailed the predictive validities of a variety of hiring tools:
  1. Assessment centres - potential (0.53)
  2. Ability tests - job performance and training (0.50)
  3. Structured interviews (0.44)
  4. Bio-data (0.37)
  5. Assessment centres - performance (0.36)
  6. Personality tests (0.33)
  7. Unstructured interviews (0.33)
  8. References (0.17)
  9. Self-assessments (0.15)

All pretty poor, really, at predicting success - performance - on the job! It amazes me that we still use references, if they are less useful than 1 in 5 of being accurate. In fact, why on earth we use anything from Bio-data on down is pretty moot. CVs don't even get a rating.

But what really surprised me was the follow up list that Anna posted; her "what is most popular" hiring assessment tools with employers (in decreasing order of popularity, based on some research Anna did between 1991 and 2006):
  1. References - 93% (predictive validity 0.17)
  2. Structured panel interviews - 88% (predictive validity 0.44)
  3. Structured one-to-one interviews - 85% (predictive validity 0.44)
  4. Competency-based interviews - 85%
  5. Ability tests - 75% (predictive validity 0.50)
  6. CVs - 74%
  7. Personality questionnaires - 60% (predictive validity 0.33)
  8. Assessment centres - 48% (predictive validity 0.53 or 0.36)
  9. Online selection tests - 25% (predictive validity 0.15)
  10. Bio-data - 7% (predictive validity 0.37)
If Anna's data is accurate, then why do employers and recruitment agencies still request the same old materials and hire as they do? It beggars belief.



Sam

No comments :

Post a Comment

Thanks for your feedback. The elves will post it shortly.