The advantages and the usability of the Expected A Posteriori (EAP) score inside the Item Response Theory: An example through the Risky Loot Box Index.

The advantages and the usability of the Expected A Posteriori (EAP) score inside the Item Response Theory: An example through the Risky Loot Box Use. Item Response Theory (IRT) models represent a psychometric approach that measures the latent trait (θ) by quantifying the probability of responding to the test items according to the level of the trait possessed by the respondent and the characteristics of the items that compose the test. Thus, this approach allows a more sophisticated assessment compared to the Classical Theory of Tests (CTT) as the scaling is centered both on the respondent and on the psychometric properties of the items. In this presentation, we will deal with the contribution of these models in the estimation of θ through the expected a posteriori (EAP) score (Bock & Mislevy, 1982), which allows to estimate θ taking into account the response patterns of the subjects at the test, and not the total test score, as in the CTT. Moreover, in order to overcome the obstacles linked to the dissemination and usability of the EAP among practitioners, we developed a digital interface by using the Shiny library in R Studio. We will use the Risky Loot Box Use (RLI; Brooks and Clark, 2019), a brief scale that measures the risky use of the Loot Boxes, as a tool to analyze the EAP properties and to apply the developed digital application.