Are Psychometric Assessments Accurate?

One of the common debates in the assessment field revolves around the accuracy of psychometric tests. Do they measure the right constructs in the right way?

The Short Answer is Yes - Psychometric Assessments are Accurate

People are complicated and the areas assessments seek to measure are also complicated. This amount of complexity gives rise to one of the most commonly held misconceptions about psychometric assessments - that they are inaccurate or unfair.

Below, we’ve outlined some of the primary reasons why psychometric assessments can be trusted to provide scientific, valuable and objective information:

1. The Primary Underlying Construct is Stable over Time

When assessing an individual’s underlying potential, the primary construct that we at Lumenii focus on is personality. This construct has been the subject of much research over the past century, and has been found to be stable over time. This means that our predictions of future performance are based on a stable, unchanging source of information.

2. Individuals Provide Insight into Themselves

Another important aspect to consider about personality-based assessments is that the individual provides insight into him/herself. It is not a maximal performance test whereby there are right or wrong answers. An individual simply provides a reflection of how he/she is likely to behave in different situations. Questions are designed to be neutral in tone to allow the individual to respond as honestly as possible.

3. Social Desirability does Happen - But it is Controlled for

While we rely on an individual to provide the most accurate information into elements that we cannot observe directly, there are sometimes cases where an individual exaggerates his/her responses. These exaggerations can be conscious (demonstrating impression management in order to appear more suitable for a role) or unconscious (not being aware of one’s own behaviours). Both of these are measured and managed within the Lumenii personality measure.

4. Results are Based on Business-Relevant Information

Typically, personality assessments provide the test administrator with a lengthy report reflecting all aspects of an individual’s traits and behaviours. While this is a detailed source of information, it does not always provide clarity or business-relevant insights. In fact, the line manager may be side-tracked by irrelevant details.

It is important to only consider those personality results relevant to the role and the company’s culture. That is why we profile a job role in terms of the 6-8 competencies essential for future performance and provide results based on these competencies. This has a great number of benefits, including:

• The individual is only scored on those competencies that are relevant for the role, and irrelevant personality scores are not evaluated. This eliminates a potential source of bias in the decision making process.

• Line managers obtain more business-specific and job-specific information, making it far easier to interpret the results and understand their likely impact in practice.

• It is more simple to understand the strengths and development areas of different individuals by looking at the competency-based scores and how they relate to the job role.

• The process of mapping personality scores to competency scores is based on more than a decade of research, with a strong predictive relationship between competency potential and competency performance.

• The competency scores provide an objective source of information. Compare this to the subjective and manual process of trying to group together personality scores to determine suitability for a role and eventual performance.

5. The Quality of Norm Groups

Participants are often concerned that their results are taken out of context. This is why it is always crucial to compare an individual’s raw scores to those of his/her peers. The Lumenii norm groups, against which raw results are interpreted, reflect the demographic breakdown of the South African general population, ensuring a fair and unbiased comparison.

6. Reliability and Validity

All tests need to undergo statistical analysis in order to prove that they are reliable (consistent over time) and valid (measuring the correct constructs in the correct manner). There are a wealth of established methodologies that are utilised in order to validate an assessment, and provide a strong, rigorous measure of its statistical performance. Based on these results, improvements are made to the test itself during the development process until it is satisfactory and able to be used. Analyses are performed on an on-going basis and additional changes made (if necessary) to continuously measure and monitor the accuracy of the assessment data.