Post by account_disabled on Mar 13, 2024 2:04:01 GMT -5
One of the most important goals of cognitive ability testing is to hire without bias. But what happens if the test is based on bias? In 2011, Berry, Clarke, and McClure found that there are subgroup differences on some tests of cognitive ability. According to their research, the scores of white participants are more related to job performance than those of black or Hispanic participants. In other words, the tests analyzed unintentionally favored white candidates. This is known as adverse impact. According to the Society for Human Resource Management, adverse or disparate impact “occurs when a decision, practice, or policy temic bias creeps into standardized testing. For example, there is clear evidence that black students in the United States score lower than whites on standardized tests. As if that were not enough, Steele and Aronson showed that black people in the United States perform worse on tests, simply by reminding them of the stereotype that they are expected to perform worse on exams than white students. And it's not just a matter of race. In a 1999 study , women's math skills were tested when they were reminded of negative stereotypes.
Women who were reminded of the stereotype that women are worse at math than men just before taking the test performed worse than the control group. Although we do not analyze in depth all the social and psychological mechanisms underlying these effects, we must be careful when using the tests, because they may favor one subgroup over another. What to do: Measure the adverse impact What measures can we take to avoid generating an adverse impact or perpetuating prejudices? First, it uses cognitive ability tests as another component of the process. If you have many selection criteria, there DM Databases will be less chance of introducing bias into the system. Secondly, pay close attention to the communications you send as part of your evaluation processes. Make sure you do not unintentionally raise any stereotypes or use sexist language that could influence test results. Finally, the adverse impact can be measured. Below, you'll find the four-step process recommended by the Society for Human Resource Management : Calculate the selection rate for each subgroup of applicants (race, gender, or other category): divide the number of candidates who pass the cognitive ability test by the total number of applicants in the group.
Determine the group with the highest selection rate . This group will be your reference point. Calculate the impact ratio: compare the highest group with the rest of the subgroups. Divide the selection rate of each subgroup by the selection rate of the highest group. Determines if any group is below the 80% highest selection rate. If yes, there may be an adverse impact. For example, the selection rates in the table below indicate that twice as many Caucasian candidates are hired compared to Latinos, indicating an adverse impact. Fuente: Society for Human Resource Management If you see an adverse ime many tools available to implement cognitive ability testing. TestGorilla cognitive ability tests At TestGorilla , we have a large catalog of tests for candidate selection . In addition to offering cognitive ability tests, we have tests in the following categories: Culture fit assessment (we actually call them “culture contribution” assessments ) personality tests Language tests Situational judgment tests Job-specific skill tests With TestGorilla, it is possible to design assessments for certain positions, as you can select or create the most relevant tests for the position.