When using a psychological test in any setting, it is of great interest for both the administrator and the respondent that the test is actually measuring what it claims to measure. This is what construct validity is all about.

A construct is a theoretical and statistical concept. It is the 'hidden truth' about individuals that we assume causes the responses to the questions in a test. With this assumption, we can use the test results to calculate estimates of the construct.

There are several ways to collect evidence for the construct validity of a psychological test. The most common method is to collect data from two tests that are designed to measure the same or similar constructs and calculate the correlation between the scores of the two tests. This is often referred to as the convergent validity of a test.

The hypothesis is that if one and the same construct causes the results of two separate tests, then the scores of the tests should correlate highly. According to European standards from EFPA, convergent validity coefficients above 0.75 are deemed 'Excellent'.

Below are results from a study Alva Labs conducted where 148 US participants completed both Alva's personality test and the IPIP-NEO-120. Both tests measure the Big 5 factors of personality. The convergent validity is 'Excellent' for all factors.

Table 1: Convergent validity for personality factors, N=148

Note that Emotional Stability is the opposite of Neuroticism, which is the factor measured in the IPIP-NEO-120. The correlation between Emotional Stability, as measured by Alva's personality test, and Neuroticism, as measured by IPIP-NEO-120, is -0.92. Reported here is the correlation after inverting Neuroticism scores from the IPIP-NEO-120.

Did this answer your question?