How to interpret psychometric test results

- some guiding principles

Kajsa Asplund avatar
Written by Kajsa Asplund
Updated over a week ago

Once your candidates have taken Alva’s personality and/or logic test, it is time to look at the results. In this article, you will find a few general principles for how to interpret the results. 

1. The amount of interpretation needed depends on how you are using the tests. If you have a fairly large number of candidates, and you use the tests as a screening tool, the need for interpretation is limited at the first stage. As long as you have set up a valid test profile (e.g. by the help of Alva’s default profiles), you can use the candidate ranking as a tool for cutoff. For instance, you may want to continue with only those candidates that have excellent or good role fit. In this case, there is really no need for you to go in and interpret the results - the test profile and candidate ranking do this for you. 

On the other hand, if you have a process with relatively few candidates, or if you use the tests later in the recruitment process, you will probably want to go in and do a qualitative interpretation of the results. Also, even if you used screening at the beginning of the process, you may want to return to the test results once you have a few end candidates to choose from and probe deeper into their profiles. 

2. Test results are strong indications rather than absolute truths. We should always keep in mind that psychometric tests are meant to capture characteristics that are not readily observable, and are quite difficult to measure. This means that there is always a margin of error. On the other hand, though, researchers have spent many decades refining the measurement of personality and logical ability. Today, the measurement quality of the tests is high. Hence, the overall results are very much reliable. Hence, it is highly unlikely that the score is way off the candidate’s ‘true’ score. In total, the concrete pointer is this: When interpreting the result, you can assume that the general tendencies are correct, but also be aware that there might be other things than the person’s true score affecting the results. 

3. Scores towards the ends of the scale indicate more uncommon qualities. Like almost all human characteristics, personality and logical ability are normally distributed. This means that most people fall in the middle of the scale, around 5-6, while results towards the ends of the scale are increasingly uncommon. A result in the mid-range of the scale can hence be interpreted as the candidate being ‘like most other people’ on this particular characteristic. A result towards either end of the scale indicates that this trait is likely to stand out. For an individual with an extraversion score of 5, for instance, this characteristic is likely not perceived as defining that person’s character. For a person with a score of 2, on the other hand, their introversion is likely to be something that other people notice and regard as typical of them.

4. For the logic test, higher is usually better. Alva’s logic test is a so-called capacity test, meaning that it measures an ability to perform a certain type of tasks. This makes the interpretation rather straight-forward: In most contexts, a higher result is better. You can read more about the definition of logical ability here, and how to interpret a low, average, and high result here.

However, in some situations, you may also choose to adopt more of a 'threshold view' on logic: For some roles, you may decide that the requirement is merely for the candidate to have a logical ability above a certain set point. If you are using Alva's test profiles, this is what a broader range represents. In these cases, the candidate will get a maximum fit score for any result that falls within the given range: Higher is only better up to a certain point. This stance is most commonly used for roles with somewhat lower complexity. 

5. For the personality test, ‘better’ depends on the profile. While Alva’s logic test measures a capacity, Alva’s personality test measures a typical tendency, i.e., the patterns of thought and behaviour that the candidate usually displays. For this type of test, there is no ‘one right answer’: What represents a good result depends on the profile. Usually, Alva’s test profiles do not define desired values for all factors and facets. This means that two candidates that both have very good role fit might still have quite different values on those facets that do not form part of the profile. Hence, two candidates might have the same probability for performing well, but would bring quite different things to the team. 

Did this answer your question?