Test reliability is important but validity is better
Valid and well-developed personality tests, when applied optimally and combined with an accurate diagnosis of the job requirements, can greatly improve the odds new hires will be stars.
However, there are still many organizations that fail to utilize this powerful tool to enhance the quality of the workforce. Others that use personality testing may not see the full benefit of these assessments — falling victim to false claims and “flavour of the month” tests and practices that fail to deliver.
The following are things to keep in mind, regardless of business context or type of test, to help organizations get the most from personality testing.
The first step, which is often overlooked, is to determine what defines success in a particular job.
There is a troubling propensity for HR managers to simply choose an off-the-shelf solution instead of first giving serious consideration to the types of characteristics a particular job calls for. The selection industry has seen the mass commoditization of personality tests and there appears to be a legion of easy-to-administer tests that claim to assess the core drivers of performance.
However, the most valid — and most defensible — approach is to use an appropriate job analysis or job-profiling technique to conduct a thorough examination of the requirements of a target position. Once the attributes necessary for success are established (such as personal drive, collaboration or boldness), a suitable test can then be selected through a careful mapping exercise.
Reliable and valid test
Once the job analysis is complete, choose a reliable and valid test to measure the job requirements. The worst mistake an organization can make is to simply assume a particular test is reliable and valid for its selection purposes.
It is critical to remember personality tests are not created equal. HR professionals should examine several tests before making a final selection. Once several tests have been identified, the HR professional should get the tests’ technical manuals and other pertinent information. When reviewing this information, keep the following in mind:
Reliability: In technical manuals, reliability refers to how consistent or stable the test scores are. Reliability can be as high as one and as low as zero. Higher values are generally preferred. In most cases, if reliability is lower than 0.6, consider a different test.
Validity: While there are several ways to measure a test’s validity, “criterion-related” evidence of validity is the most compelling — and the most difficult to obtain. Criterion-related evidence of validity shows those who score higher on a test tend to receive higher performance ratings. Optimally, HR professionals should seek out criterion-related validity evidence that shows correlations between test scores and job performance within a similar role or job family. However, some personality traits, such as conscientiousness, have been shown to relate to performance in a range of positions. For these traits, a “validity generalization” argument can be used as a legally defensible substitute for evidence of a test’s validity for a particular job. Generally speaking, evidence of validity is more important than evidence of reliability, but harder to obtain. Consequently, more than a few tests have been marketed on the basis of strong reliability while forgoing evidence of validity. Indeed, the fact a test consistently measures the same thing does not mean it will be helpful in identifying high-performing employees.
Resistance to faking: Some personality tests are susceptible to applicants deliberately distorting (or faking) their responses in a manner that may make their scores less valid. For example, if the test includes a question such as “I am a hard worker (true or false),” even the laziest of applicants might choose “true” in an effort to increase his chance of being hired. There has been a vast amount of research aimed at managing the effects of faking on personality tests. An assessment expert’s input can help HR professionals choose a test that research shows to be “fake-resistant.”
Test drive the test
Once an HR professional has selected a reliable and valid test, she should “test drive” the assessment herself. Things to keep in mind include: the accuracy of the suggested time limit; the nature of the target job; the applicant pool; and the test’s “face validity” — that is, whether test-takers can see the direct connection between the questions being asked and the content of the job in question. Organizations have pulled valid and reliable tests from use due to test-takers being frustrated by seemingly irrelevant items and test topics.
Following the principles described here requires an investment of time and effort, but research shows the result — a successful employee — will pay for the time and effort 10 times over. Personality testing can give organizations a better idea of what a candidate will do versus what he says he can do, and provide real business impact for an organization.
Henryk Krajewski is vice-president and national practice leader of talent management at Right Management in Toronto. He can be reached at (416) 926-1324 ext. 249 or [email protected]. Richard Goffin is a professor of industrial and organizational psychology at the University of Western Ontario in London, Ont. He can be reached at (519) 661-2111 ext. 84641 or [email protected].