Diversity practice is based on three principles. First, there is the recognition that the skills and aptitudes required by employer organizations can be found in people from diverse backgrounds across the society.
The second principle is that the demonstration of those talents, skills and aptitudes is likely to be influenced by those differences in backgrounds. Two people with equivalent intellectual abilities, but with very different backgrounds and personal histories may display very different linguistic and behavioural styles.
Therefore the third principle is that employers need to ensure that all HR procedures are designed to avoid bias and provide equality of opportunity.
The potential effects of diversity present a particular challenge when organizations need to assess the suitability of job applicants from a range of social backgrounds. These assessments need to evaluate their potential to meet the requirements of the jobs, while not being corrupted by irrelevant individual differences.
Many employers have opted for various types of psychometric testing in the belief that they provide the most ‘objective’ assessments. But how objective or ‘culture-fair’ are assessments using psychometric tests?
On closer inspection, we find that some widely used tests result in significantly different average scores for white and black candidates. Using these tests, recruiters systematically exclude black candidates, concluding that “they just aren’t up to the job”. This assumes that the test scores are equally valid as measures of the abilities of the two groups and that, as predictors of subsequent performance, the scores are equally accurate.
Unfortunately, the test users may be wrong on both assumptions. It is essential therefore that any chosen test is equally suitable for a diverse group of applicants. Tests used in selection should show no evidence of bias or ‘differential validity’ between groups.
Even where an appropriate test has been chosen, the conditions under which it is administered could bias the results.
It has been shown that where a candidate has a fear of being assessed against a negative stereotype, this fear (called ‘stereotype threat’) can depress their performance. In addition to evidence of this effect where black students fear being stereotyped as poor performers, it can affect women where they are thought to be inferior to men or students from poorer backgrounds facing negative stereotypes about their intellectual abilities.
In a series of experiments where students of equivalent abilities were assigned to different test conditions, it has been shown that where stereotype threat is aroused for some students, their performance was significantly below that of the students who did not feel that threat. The performance of test candidates has been shown to be influenced by how the tests are introduced and even the choice of test administrators.
To further complicate matters, we also know that some measures of school achievement vary in their ability to predict subsequent performance. For instance, on average, students from comprehensive schools are likely to achieve higher degree classes at university than students with similar attainment from grammar and independent schools[1]
In designing selection exercises, employers need to be aware of the ways in which social and cultural diversity can affect the processes and ensure that they ‘diversity-proof’ the procedures that they choose.
Candidates can also take steps to improve their test performance as previous experience and familiarity with tests of a similar type can affect test scores. For best results, test candidates should take advantage of information provided by recruiters about the test procedures to be employed. The can also look out for the variety of books offering opportunities to practise in preparation for selection exercises in recognition of the positive effects of familiarity.
For fair selection – as a minimum – recruiters need to take diversity into account in
- the training of recruitment personnel;
- the choice of tests and exercises;
- how the test situation is organised;
- how assessments are introduced;
- the choice of people to administer the assessment procedures; and
- signalling that the organization welcomes and values diversity.
Candidates should
- find out all that they can about each employers’ selection procedures;
- look out for information on psychometric tests published in newspapers, books and magazines; and
- take every opportunity to practise the different types of reasoning or aptitude tests that they may meet in recruitment exercises.
© Dr Marie Stewart MBE
[1] Sutton Trust (2010)
Tags: assessment, bias, diversity, psychometric tests, recruitment
Living in a Diverse World 