This research represents an unforeseen outcome of the authors' National Science Foundation Innovation Technology Experiences for Students and Teachers (ITEST) program grant in science education. The grant itself focused on the use of serious educational games (SEGs) in the science classroom, both during and after school, to teach science content and affect student perceptions of science and technology. This study consists of a Bayesian artificial neural network analysis, using the preintervention measures of affect, interest, personality, and cognitive ability, in members of both the treatment and comparison groups to generate the probabilities that students would opt into the treatment group or choose not to participate. It appears, from this sample and the sampling methods of other related studies within the field, that despite sometimes profound results from technology interventions in science, interventions are affecting only those who already have a strong interest in STEM due to the manner in which participants are recruited.