Individuals who possess extensive domain knowledge use their knowledge when understanding, discussing, and remembering events. The purpose of this study was to assess the extent to which natural language processing (NLP) tools could be used to predict domain knowledge from typed descriptions of events. Participants watched videos of basketball and recalled them after viewing. Knowledge of basketball was assessed. NLP tools were utilized to assess whether linguistic features of participants’ event descriptions could be used to predict domain knowledge. Moreover, the extent to which linguistic indices could be classified as relating to linguistic complexity or features of mental model construction was explored. Results from machine learning models suggest that domain knowledge (high, low) could be predicted with up to 90% accuracy. Additionally, 90% of predictors could be categorized. Higher knowledge individuals tended to describe events with more linguistic complexity and produced more words related to spatial, temporal, and social relations.

Researchers have long been interested in understanding and closing the “reading gap” that exists between White and racially marginalized students. This study explored whether group differences in inference strategies (i.e., bridging and elaboration) and comprehension performance existed among college readers. Three hundred college participants who self-identified as White, Black, or Hispanic completed a think-aloud task along with measures of reading proficiency and comprehension. Results from hierarchical regression models indicated that group differences in elaborative strategies were present, but differences in bridging strategies and comprehension performance disappeared when foundational skills in reading were included in the models. The results are explained in terms of inequities in educational experiences prior to entering college.

The current study examined the extent to which the cohesion detected in readers’ constructed responses to multiple documents was predictive of persuasive, source-based essay quality. Participants (N=95) completed multiple-documents reading tasks wherein they were prompted to think-aloud, self-explain, or evaluate the sources while reading a set of four texts. They were then asked to write a source-based essay based on their reading. Natural Language Processing techniques were used to automatically analyze the cohesion of the constructed responses at both within- and across-documents levels. Results indicated that within-document cohesion was negatively related to essay quality, whereas across-documents cohesion was positively related to essay quality. Further, these relations differed by instructional condition such that strategic instructions to either self-explain or evaluate sources seemed to promote across-text integration, compared to thinking aloud. Overall, this study indicates that the cohesion of constructed responses to text can provide insights into the coherence of the mental representations readers construct while reading multiple documents.