The current study examined the construct validity of the Situational Test of Emotional Management (STEM), a newly developed measure of emotional intelligence. A logistic regression approach was utilized to assess the STEM, at the item-level, on two factors, gender and Extraversion. It was hypothesized a logistic model including gender and Extraversion would significantly predict classification of participants into two groups (correct response on STEM items and incorrect response on STEM items) beyond a simple intercept-only logistic regression. Data analysis revealed that the gender and Extraversion did not significantly enhance the classification rates of participants into the two groups, but gender was a significant univariate influence on four items. Explanations as to why the hypothesis was not fully supported are discussed, as well as avenues for future research.