Validity
Respected researchers in organizational development such as Lyle and Signe Spencer of McBer and Company estimate that the unstructured interview assessment and selection process used by more than 95% of organizations worldwide has a validity correlation with job performance in the range of .05 to .19. In contrast, the validity correlations for structured behavioral interviews are estimated to be in the range of .48 to .61.
One objective of the Optimum Performance Profile is to push these validity correlations for structured behavioral interviews even higher by improving the information available to interviewers and by focusing the questions asked in these interviews on those behaviors and competencies essential to success in each job role.
Even as a stand alone assessment the validity correlation for our assessment instrument is very strong. Correlations of r= .30 to .39 are considered strong, and the correlations for our instrument to performance in each of the five job roles range from .40 to over .50 with over a 95% confidence level that the results were not due to chance. Item reliability for the individual behavior trait factors in our instrument is also very high, ranging from .7 to .9 and above for all behavior traits with each trait encompassing multi-item scales of 3-11 items each.
Our research demonstrates that no single competency and no single component of the selection process has adequate predictive validity to stand alone as a selection tool. Since no one selection tool can provide a complete assessment of a candidate’s ability to use the critical competencies required for each job role, the best way to increase predictive validity is to conduct multiple assessments based on a validated and integrated scorecard for each job role.
In Competency at Work, Lyle and Signe Spencer report the following criterion validity correlations with performance for various methods of assessment:
Assessment Method Correlation
- Assessment Centers .65
- Interviews (behavioral) .48-.61
- Work sample tests .54
- Ability tests .53
- Personality tests .39
- References .23
- Interviews (non-behavioral) .05-.19
We conclude that selection should be viewed as a process that runs on a continuum from continuous and proactive recruiting through screening, testing, interviewing and orientation.