In a practical sense, the most valuable attribute of a selection procedure (i.e., personnel assessment method) is the degree to which it successfully predicts future job performance, job-related learning, and other criteria. The term that describes the ability of an assessment tool to predict future performance is called predictive validity. The greater the predictive validity of a selection procedure (or some combination of assessment procedures), the better it is at predicting the outcomes described above.
Selection procedures with high predictive validities also have more value for organizations, via increased productivity, output, and learning ability of their workforce. In a seminal article, Schmidt and Hunter (1998) conducted a meta-analysis (or statistical combination) of thousands of studies over 85 years to determine the predictive validity of 19 different selection procedures, both individually and in combination with general mental ability (GMA).
GENERAL INTELLIGENCE COMES OUT ON TOP
The authors found GMA (also known as g, general intelligence, and general cognitive ability) to be the single best tool for selection. GMA measures have numerous benefits, including:
- low-cost
- highly predictive of job and training performance
- supported by almost a century of research
- predict performance for all types of jobs and at all job levels
- not reliant on applicants having previous experience (as is the case with work sample and job knowledge tests)
The authors even noted that GMA can be “considered the primary personnel measure for hiring decisions, and one can consider the remaining 18 personnel measures as supplements to GMA.” Therefore, according to their findings, the important question in selection is not “what single procedure should we use?” but rather “what procedure should be combined with a GMA assessment?”
EMPLOYMENT TESTING BEYOND INTELLIGENCE
Their meta-analysis found that the three combinations with the highest predictive validity (in decreasing order) were: GMA plus an integrity test, GMA plus a structured interview, and GMA plus a work sample test. The first two can be used with entry-level applicants and experienced employees, while the latter can only be used with employees who have previous job knowledge or experience.
An integrity test measures conscientiousness as well as aspects of agreeableness and emotional stability. A structured interview has carefully prepared, precisely scored questions based on an analysis of the open position. Finally, work sample tests are composed of real-life simulations of the work to be performed by applicants. In terms of cost, integrity tests are by far the most affordable (they come in the form of questionnaires), structured interviews can take time to develop and use (but they are much more valid than unstructured interviews), and work sample tests are the most costly of the three. However these are still likely the best approach when hiring for positions that need specialized skills.
POOR PREDICTORS OF EMPLOYEE PERFORMANCE
It is also important to note the measures that do not predict future performance. The authors found that many measures have little or no predictive validity whatsoever. Age and graphology (the psychological analysis of handwriting) were found to have no predictive validity, while personal interests, years of education (although this predicts training performance fairly well), the T&E point method, and job experience (years in a similar job) had the lowest predictive validities for job performance.
PRACTICAL IMPLICATIONS FOR ORGANIZATIONS
When it comes to designing an effective selection procedure that will help your organization hire the right person, the authors recommend using either GMA plus an integrity test or GMA plus a structured interview. Both approaches are valid and inexpensive ways of predicting how applicants will perform both in their jobs and during their training. As the authors note, “The gains from increasing the validity of hiring methods can amount over time to literally millions of dollars.” But keep this in mind too – using ineffective methods can have the opposite effect and cost millions of dollars in lost productivity.