The purpose of this article is to clarify possible doubts regarding the transition from the fluid reasoning test (FR) to the adaptive reasoning test (AR). The AR test was validated against the RF test, and we explore this process in more detail below.

It is important to highlight that, when comparing the application of the AR test with the RF test, we need to take into account variations in external factors (such as conditions in which the person is performing the test and the environment that surrounds them). These variations change the scores by a certain sample amount – what we call an “error” in the individual's score. Thus, even if we applied the same test twice with the same sample of people, the results obtained would certainly be different (although close).

About the RA test, we have that it presents an r correlation of 0.81 with the RF test. Given the high correlation and the “error” associated with any measurement of grade, it is possible to say that the tests measure the same attribute (in this case, the cognitive ability of the respondent). In addition, since the two tests have the same rating scale (0 to 100), it is possible to verify that they are close, as explored in the graph below:

Only a few notes show a large discrepancy between both tests, which reinforces the similarity between the two models. We can also see that the RA and RF test grade distributions are similar, as seen in the graphs below:

In the end, we highlight that the AR test was applied with three different samples and showed similar performance to the RF test in the three applications. This exercise ensures the quality of the retest and ensures a reliable transition in the system.