In the United States, the greatest decline in the number of students in the STEM education pipeline occurs at the university level, where students, who were initially interested in STEM fields, drop-out or move on to other interests. It has been reported that “of the 23 most commonly cited reasons for switching out of STEM, all but 7 had something to do with the pedagogical experience.” Thus, understanding the characteristics of the pedagogical experience that impact students' interest in STEM is of great importance to the academic community. This work tests the hypothesis that there exists a correlation between the semantic structure of lecture content and students' affective states. Knowledge gained from testing this hypothesis will inform educators of the specific semantic structure of lecture content that enhance students' affective states and interest in course content, toward the goal of increasing STEM retention rates and overall positive experiences in STEM majors. A case study involving a series of science and engineering based digital content is used to create a semantic network and demonstrate the implications of the methodology. The results reveal that affective states such as engagement and boredom are consistently strongly correlated to the semantic network metrics outlined in the paper, while the affective state of confusion is weakly correlated with the same semantic network metrics. The results reveal semantic network relationships that are generalizable across the different textually derived information sources explored. These semantic network relationships can be explored by researchers trying to optimize their message structure in order to have its intended effect.