The popularity of behavioral testing presents an interesting set of challenges to those concerned with test cheating, especially with regard to pre-employment testing delivered over the Internet (see https://crm.hr.com/servlets/sfs?t=/blogs/blog.show&e=UTF-8&i=1116423256281&l=0&blogid=1204887700655). For unlike objective tests (i.e., tests on knowledge, skills or cognitive abilities, most of which include questions with right and wrong answers), cheating on a behavioral test does not require crib notes or sneak peeks at questions, just the willingness to “shade” the truth.

This is because most behavioral or psychological tests take the form of self surveys where job applicants are asked to respond to a series of questions or statements that correspond to various personality factors. For example, a candidate might be asked to specify their level of agreement with a statement such as “Constant interruptions from customers and fellow employees can significantly diminish workplace productivity.” If the test is being given for a customer service position, a high level of agreement with such a statement might be scored low since customer service requires a high degree of interaction with clients.

“Cheating” on these types of tests (also referred to as “gaming” or “faking”) consists of selecting responses based on which answer the test taker thinks would show him or her in the best light (vs. a response that most accurately reflects the candidate’s honest opinions). Professional test developers are divided on the level of faking that takes place on behavioral tests and whether faking correlates with other factors such as cognitive abilities. Research has shown that even high levels of faking do not diminish overall test validity, but employers still need a way to determine the level of frankness or candidness of individual test candidates taking a behavioral assessment.

Faking has clear parallels with other aspects of the screening process. Resumes and interviews are all subject to exaggeration and, in increasing cases, outright fraud. Unlike these other screening tools, however, behavioral assessments can be built to automatically detect potential “faking” responses.

Most professional behavioral tests include some kind of candidness measure, i.e., questions or sets of questions designed to identify those who might not be answering questions honestly. These questions are used to create a specific score designed to inform test administrators of detected levels of potential faking. Research has also shown that test questions that are written in a non-invasive style (like the sample question above that is not explicitly asking someone to talk about him or herself) tend to elicit more honest responses.

Some test designers suggest that test instructions should include warnings of the consequences of detected faking and request honesty. However, the negativity and suspiciousness that this approach may trigger in the applicant could outweigh the intended benefits of this tactic.

Instructions may also ask candidates to answer quickly, with the first thought that comes to mind. This approach is based on the premise that it takes time to lie. But including this in the instructions offers no assurance that the applicant will comply.

Reducing the impact of faking boils down to making every possible attempt at increasing test validity through best practices in test development. Incorporating work-related (context-specific) instructions and behavioral items, and less invasive attitudinal test questions into the assessments are critical in increasing validity, increasing the applicants’ comfort level and reducing legal challenges.
Find out more about behavioral testing at http://www.fadvassessments.com/product_view.pl/aptitude.