top of page
The Solution
Synthetic Validity is considered an answer to this problem of validating assessments that cause adverse impact, but have high predictive power, so that businesses can use these assessments even for positions where they hire less than 100 employees, without having to worry about legal liability.
Synthetic validity is a form of validity generalization, in which jobs are characterized by a unique set of common components. By determining the validity of predictor tests (such as cognitive abilities and work personality characteristics) for each of these components, it is possible to estimate the validity of a set of predictor tests for any particular job.
In collaboration with the world’s foremost experts in synthetic validity, RightHire has implemented a novel solution to this second challenge without the need to collect validation data for each O*NET occupation. This method involves using I/O psychologists as subject matter experts (SMEs) to estimate the validities of the 27 O*NET GWAs and 11 Work Context items for RightHire’s battery of 17 cognitive and 16 work personality tests. This approach greatly improves the practicality, ease, and speed of building a viable and scalable employee selection system.
The second challenge has been more difficult to solve because to estimate the validity of a predictor test for each component requires gathering validity data between the predictor test and each of the more than 950 O*NET occupations. As a result, commercial implementation of a scalable synthetic validity solution has not been traditionally feasible.
The challenges to implementing synthetic validity are first in defining the job component taxonomy that can model any job and second in establishing the validities of predictor tests of each of these job components. In recent years a scalable solution to the first challenge has been offered through the U.S. Department of Labor’s O*NET database which can be used to model any job as a combination of its 27 generalized work activity (GWA) and 11 Work Context components.
Technology
Technological Breakthrough
See What Makes our Platform Tower Over the Competition
.png)
254 Chapman Rd, Ste 208 #18212, Newark, Delaware 19702
Info@righthire.com
+1 (877) 285-3968
Socials
Connect Socially and send us a Message
Socials
Connect Socially and send us a Message
254 Chapman Rd, Ste 208 #18212, Newark, Delaware 19702
Info@righthire.com
+1 (877) 285-3968
The Problem
The Uniform Guidelines on Employee Selection Procedures (UGESP), established by the Equal Employment Opportunity Commission (EEOC) along with other federal agencies, state that if pre-hire assessments cause an adverse impact on protected groups, they must be validated to ensure that they are job-related and consistent with business necessity.
However, the problem is that small businesses rarely hire at least 100 employees in the same job type. Even large companies don’t hire at least 100 employees in most of their positions, except for positions such as customer service representative. This is the reason why even large companies typically use pre-hire assessments only for positions where they hire in volume and can therefore validate the pre-hire assessments for such positions. Small businesses don’t use assessments often because the assessments available in the market are generic in nature and not tailored to specific positions.
Assessments that include cognitive ability tests generally cause adverse impact, but they are one of the most highly predictive tests available in the market. Traditionally, in order to validate an assessment that consists of cognitive abilities and work personality characteristics (such as RightHire’s tests), at least 100 employees who have taken the assessment for a particular type of job (such as a customer service representative) should be hired, their job performance say after a year needs to be evaluated (for example, through supervisory ratings of job performance), and the correlation between the hired employees’ assessment scores and the job performance evaluations needs to be calculated. If this correlation coefficient is at least 0.2, the assessment can be considered useful to predict job performance and maybe legally defensible.
The industry-standard Principles for the Validation and Use of Personnel Selection Procedures produced by the Society of Industrial-Organizational Psychology (SIOP) explicitly identifies synthetic validity as a viable solution. Two court cases to date involving selection procedures based on synthetic validity were decided in favor of the synthetic validity approach, though these were summary judgments with bounded legal precedent. Although future legal challenges will further clarify the issue of defensibility, peer-reviewed expert opinion is that synthetic validity will be considered a sound foundation for employee selection systems.
.png)
.png)
bottom of page