“What is the best pre-employment assessment to use for hiring?”

I receive this question weekly. My response: NEVER EVER, NEVER EVER, NEVER EVER, and in case I am unclear, NEVER EVER use an assessment tool as the primary or only data for a hiring decision. I often get, “Well the publisher shows in their manual that the ability to predict best hires in 85%” or “We’ve been using XYZ and we think people are not gaming it because there is much on the web about it, so what compares?”

Measurement Error

Why is my response so strong? Measurement error. Every psychological tool, and pre-employment assessment created, has measurement error. That means there is a “true score” and a “report score” and the difference is the amount of error there is in the assessment. And there are lots of reasons for error; the outcomes however of ignoring it is false positives (selecting those who really don’t quality) and false negatives (rejecting those who are capable). Lots of inefficient and waste occurs by ignoring measurement error and not having multiple processes in place to validate the capabilities of the individual involved.

Example: The WISC-V

Consider the measurement error of one of the most examined assessments available: WISC-V. As a measure of intelligence for children it is used for lots of purposes, including who gets into specific programs. The measurement error is 15 points. So if the cut off is 120 to get into a program and a kid scores 115, we have a problem.

A second problem is the amount of true accounting of one factor for another. So if you have a personality tool that is being used for pre-employment and various studies have shown that there is a significant correlation of .70 between Scale X and Behavior Y, then only 49% of what is measured and the target behavior share the same content. Oops.

Yes, I know 100 million people have used XYZ Assessment and 500,000 employers are satisfied users. That doesn’t mean that the use is effective. The use of the tool may reduce some anxiety, “At least we have an independent measure…..” but that is a salve for discomfort rather than evidence of efficacy of use. Scientific methods have given us a path through the dark woods of assessment, if we would only listen.

A Solution

Using a profile of Roles and Practices for a specific job helps form the use of structured questions in a methodical interview process. There are several layers to structured interviews which are designed to increase accuracy and reduce bias in selection. A FIT report for each candidate which is based on the benchmark profile (informed by your stakeholders…not provided by a vendor) contributes to the decision making. This is important because your context is unique and no vendor can fully understand its uniqueness.

If there are some very specific attributes that the organization also wishes to measure when the 3 of final few applicants are under consideration, then the use of well researched and validated personality tool relevant to the job might be of value.