Psychometric tests and questionnaires are a useful way of finding out what people are like. They are obviously not the only way of doing this as everyday we make “assessments” about those we meet. However, human judgement is not always accurate and so psychometrics provide an objective way of adding to what we know. This is particularly important in high stakes situations like hiring new employees.
In a hiring context the results from psychometrics are best used as part of the decision-making process. In particular, when the information is needed for some form of selection. This is when candidates are classified according to the presence or amount of specific work-related attributes. Bearing in mind that the same attributes -- and other qualifying criteria -- will also be checked by reviewing resumes and application forms, and assessed through interviews and other similar activities.
For tests, when results are used for screening, the information is used to select candidates “in” or “out” of a selection process. Likewise they can be used for classifying candidates, as say, “preferred,” “acceptable,” or “potential risk.” In addition, the same information can be used to match candidates against particular roles; and once people are onboarded, as an element of engagement, development, or career planning activities.
What is a “good” test?
Whether it’s a questionnaire or a test, there are five key qualities of psychometric assessments. All are important if an assessment is to be accurate, fair, and unbiased.
- Standardized. What someone is asked to do is clearly defined in both content and activity. For example, everyone gets the same instructions, the same questions, and the same answer options.
- Objective. The task and how people perform -- the output -- is independent of the influence of other people. This means how the assessment is delivered, scored, and interpreted is the same for everyone.
- Reliable. Whilst people’s performance will always vary, the normal variation is known, and so results can be judged as being inside or outside the appropropriate range for the assessment. This signals the robustness of the results.
- Scaled. The results are presented using statistically meaningful and understandable units of measurement. There are a variety of standard scales that can be used but the important thing is they all allow people to be compared directly with each other.
- Valid. The assessment measures what it claims to measure -- for example, a personality questionnaire is measuring personality and not cognitive ability -- and that what is being measured is directly relevant to the performance of a particular job.
Making the right choice
Thinking about the key qualities, it’s obvious that some types of assessment are more “psychometric” than others! For instance, unless interviews are very tightly scripted, scored, and rated, the results are likely to be far less objective than a psychometric assessment.
The implication is that while an interview gives a “feel” for the candidate, and may for example be a useful way of asking how someone would approach a work task, it is not an effective method for assessing underlying personality, motivation, interests, or values. In addition, all these more nuanced attributes are often confounded by someone’s (good or bad) communication skills.
The same can be said for cognitive tests. If an organization wants to understand someone’s “raw” problem-solving ability, or how quickly they might learn new skills, a test is a better way of doing this than asking a series of interview questions. There’s often simply too much “noise” to really appreciate -- during an interview -- the subtleties of someone’s reasoning style and ability.
The consequence is that employers need to appreciate the quality of data that various assessments provide. The “five qualities” are a good starting point. And obviously while it’s unlikely that someone will be employed without an interview -- and in some cases, without having delivered a presentation or participated in some form of work simulation -- it’s always wise to supplement these with more objective information. This can then be used to help consistently and fairly compare one person with another.
Do your homework!
There’s a lot of weird stuff out there. As ever, when it comes to tests and questionnaires the best advice is to check the psychometric qualities of a particular product. The same advice applies to the more eccentric ways of attempting to assess people. For example, thankfully brain scanning is still some way off! However, there are two other methods that deserve a mention.
Astrology, or predicting personality based on time, location, and date of birth, has yet to be proven as a reliable and valid way of determining someone’s personality. And Graphology, or predicting someone’s personality from an analysis of their handwriting, falls into the same unproven category. This speaks volumes as both methods have been around for centuries.
However both techniques are still used in hiring. Graphology is popular in a number of European countries, and astrology also has an enthusiastic global following. The latter has a strong following in China. So whatever your point of view, the important thing is to conduct the necessary technical due diligence on any assessment you’re considering. This is the only way to ensure that candidates -- and employees -- have a fair and credible assessment experience.
Traitify builds assessments that help attract, select, and engage employees based on their personalities. If you would like to know more, connect with Traitify.