ABSTRACT. This session includes five exemplars of how Walter Kintsch's seminal work in discourse comprehension and cognitive science inspired groundbreaking research and development leading to innovations in computational modeling, educational technologies, and the neuroscience of language. The breadth of the work described in this session exemplifies how Kintsch’s drive to support interdisciplinary, collaborative science is crucial to propelling both theoretical advancements and real-world applications in the field of text and discourse processing.

ABSTRACT. Walter Kintsch's contributions profoundly impacted research on discourse comprehension and interdisciplinary cognitive science. Each of his advances produced a cascade of theoretical and empirical investigations by himself and many others, which validated those advances. Some, to emphasize a few, were the propositional representation of text, the construction-integration model, multidimensional situation models, and global memory search in text retrieval. Walter's remarkable insights regarding cognitive science inspired all those around him in their research endeavours and beyond.

ABSTRACT. The functional neuroanatomy of text comprehension has been studied in two ways: On one hand, neuropsychological patient studies are important for diagnosis and treatment of non-aphasic language disorders. On the other hand, neuroimaging methods are useful for evaluating models of comprehension and they have enhanced our understanding of discourse processing. In this talk, I reflect on the impact Walter Kintsch’s work has had on neuroscientific research and sketch some recent developments of the field.

ABSTRACT. Kintsch (1998) showed that text comprehension emerges from interactive processes between the reader and the text, leading to a reader’s Situation Model (SM). Empirical findings further reveal SM formation also draws on readers’ embodied processes. Readers exhibit elevated gesture production during inference-making (N=22); increase gesture production to compensate for SM-impoverished texts (N=48); while restricting gesture production selectively impairs inference-making but not textbase recall (N=79; N=42). These findings inform the design of educational texts and assessments.

ABSTRACT. Kintsch's approach to understanding the cognitive processes of reading comprehension incorporated cutting-edge computational modeling approaches to solve both practical and theoretical issues in cognitive science. Kintsch moved the field of discourse processing by integrating automated semantic modeling as a central core to the knowledge construction and integration process. His interdisciplinary work exemplifies the processes needed in cognitive science to continually advance theories as well as produce applications that have impacts outside of just academia.

ABSTRACT. Developers of learning technologies face endless choices about when and how to deploy digital features to support meaningful student outcomes. The Construction-Integration (CI) Model, as a general theory of comprehension focused on underlying cognitive processes, provides a framework to guide technology design and assessment in educational contexts. This talk will outline approaches to using the CI Model as a design framework, demonstrating its application to common digital interactions and resulting learner outcomes.

ABSTRACT. Audience discussion

ABSTRACT. Language comprehension is accompanied by the reactivation of experiences. To investigate how the processing of dynamic directions influences response accuracy and speed in text comprehension, participants listened to narrative texts while watching clips that matched or mismatched the directions described in the texts. No interaction between clip and story direction was found. Instead, we found a main effect for story direction, with slower and more inaccurate responses to away-stories than to towards-stories.

ABSTRACT. We present a conceptual and a direct replication of Blumenthal-Dramé (2021)’s self-paced reading study, which indicated that a connective facilitates reading more in synthetic languages (German) than in analytic languages (English). Contrary to Blumenthal-Drame (2021), we do not find any interaction between language and relation marking in our experiments. A reanalysis of the data of the original study reveals that the interaction between language and relation marking was likely a spurious effect.

ABSTRACT. The current study investigates whether inferential processes typically considered passive/automatic are influenced by task-related standards of coherence when reading short narrative texts. To manipulate standards of coherence, participants completed self-paced reading and lexical decision (LD) tasks and answered text-specific or general setting comprehension questions. Results indicate a main effect of standards of coherence: Faster LD reading times for text-specific versus general setting comprehension questions. Study findings inform theories of cognitive processing during reading comprehension.

ABSTRACT. Involuntary thoughts are a common part of everyday life. Our study explores how types of involuntary thoughts differ from one another based on language use and age. Younger and older adults (N=314) provided examples of three different types of involuntary thoughts. Using machine learning, we predicted thought type and age, above chance, based on language features alone. We also found significant differences in descriptions of thought types as a function of sentiment.

ABSTRACT. The present study found that participants with lower proficiency in their second language (L2), as opposed to their first language (L1), exhibited increased frequencies of mind-wandering and decreased comprehension performance when reading in L2 compared to L1. This result aligns with the executive control failure account and situation model view of mindless reading. The effect of reading proficiency on comprehension was, in part, explained by its influence on the occurrence of mind-wandering.

ABSTRACT. We investigated how mind wandering (MW) impacted memory for texts. Participants had to read or listen to texts interrupted after a short period or after a longer period and were asked whether they were focused on the text at the moment the interruption occurred. They subsequently performed a sentence verification task (verbatim, paraphrases, inferences, unrelated). The results showed that MW mostly altered processing of verbatim and that the deleterious effects were observed after early interruptions.

ABSTRACT. This study investigates whether prior knowledge indirectly affects intertextual comprehension performance through its effect on intratextual comprehension. Based on raw data from extant studies, it finds that prior knowledge positively affects intertextual comprehension, mediated by intratextual comprehension. This effect varies depending on whether comprehension is measured through idea reproduction in essays or idea linkage recognition across texts. These findings highlight the nuanced relationship between individual comprehension skills and the nature of comprehension assessment methods used.

ABSTRACT. The current study uses two novel measures to assess accuracy and coherence of psychology content knowledge. These measures were designed to build upon the Multidimensional Knowledge in Text Comprehension framework. Students’ psychology knowledge accuracy, psychology knowledge coherence, and science knowledge predicted 10% of the variance in psychology course grades. However, when reading skill was added to the model, it emerged as the only significant predictor, indicating the importance of reading comprehension in course performance.

ABSTRACT. This study examined the effects of summarization on multiple text comprehension. Participants read four texts about meat consumption, and were asked to summarize each text or reread. They responded to intratextual and intertextual verification items. Preliminary results demonstrate that summarizing benefitted intratextual inferencing, but not intertextual inferencing. Results suggest summarizing encourages focusing on texts as separate entities rather than intertextual integration. Connections to theories and instructional implications will be discussed.

ABSTRACT. The study examined: the interplay of task context and complexity on task model, and the impact of task model on information search strategies and outcome. University students completed search tasks framed in an academic or a non-academic context. Task context and task complexity mostly impacted search strategies, confirming that context can increase standards of coherence. Task model standards increased source recall and mediated the impact of task context on search strategies and source recall.

ABSTRACT. In this study, we identify individual differences in students’ epistemic cognition as they complete an online inquiry task in an unrestricted internet environment. To do so, we evaluate and integrate students’ think-aloud, behavioral, and post-task interview data. After identifying different profiles of students’ epistemic cognition, we plan to assess whether these profiles predict their performance on an argumentative essay task, controlling for students’ digital reading and writing ability, prior knowledge, and media usage and attitudes.

ABSTRACT. Sixty-five undergraduates completed an authentic academic search result evaluation task. Students evaluated a set of preselected texts based on each source’s usefulness and trustworthiness, and justified each evaluation. Results indicated students with stronger reading comprehension skills more appropriately evaluated less reliable information as less trustworthy, especially when they had strong pre-existing topic beliefs. Students with stronger reading comprehension skills also utilized more critical criteria during evaluation, such as assessments about evidence and source quality.

ABSTRACT. Based on theory about goals and strategies for disciplinary learning, this presentation catalogues the implementation of steps-chart activities in introductory psychology. Students’ thinking about theories and evidence were scaffolded, including classroom discussion, comparisons with peers, and completing causal models. We will assess learning at the end of the term vs the same items from a prior year and student performance on these activities. We will also present challenges and successes from the instructors’ experiences.

ABSTRACT. iSTART is a reading strategy tutor that affords generative practice opportunities to improve reading comprehension skills. This study examined whether generative practice in iSTART also supports students’ knowledge building. Undergraduates in psychology courses were given training and deliberate practice opportunities in iSTART while reading psychology texts. Finally, their reading comprehension and psychology knowledge were assessed. There was not a significant relationship between the students’ generative practice iSTART and their reading comprehension and psychology knowledge scores.

ABSTRACT. Previous research comparing reading modalities has focused on narrative texts, children, and second-language learners or those with learning disabilities. The current study examines undergraduate students’ expository text comprehension across visual and audio modalities. We also investigate how text narrativity and concreteness features, as well as participants’ prior knowledge and interest, interact with modality. We present results and discuss their implications for post-secondary education in the context of text-to-speech applications

ABSTRACT. It is not well understood if children’s engagement, both cognitive and emotional, and metacognition differs between paper books and digital devices. In this study, elementary-school children (N = 62) read from paper and digital books in a counterbalanced, within-subjects design. Engagement and learning were similar between paper and digital books. However, children reported better remembering the content from the digital book in a forced-choice comparison to paper books despite similar actual memory across medium.

ABSTRACT. This study explores the impact of video and audio feedback compared to written feedback on usefulness perceptions and academic performance. First-year teacher training students completed high-stakes assignments on educational psychology and received feedback either in video, audio, or written format. Controlling for GPA and engagement, results showed no significant differences in perceived usefulness of feedback formats, but video feedback positively impacted academic performance compared to written feedback. These results highlight potential implications for instructional design.

ABSTRACT. To test a prediction of the cultural origins hypothesis, we asked blind (n = 23) and sighted (n = 20) adults to recall the modality of words whose definitions they had learned, and to indicate their confidence in their response. We found blind participants had increased metacognitive confidence, despite being no more accurate in their judgments. The findings are consistent with the cultural origins hypothesis and support nonverbal broadcasting as a vehicle for cultural learning.

ABSTRACT. We investigated the relative and combined benefits of retrieval (vs. restudy) practice and multimedia (vs. text) learning (N = 228). The results showed a multimedia effect after 1 week, but neither an effect of practice type nor an interaction effect. Furthermore, learners were sensitive to the multimedia effect, while they predicted restudy to trump recall practice. The theoretical and practical implications will be discussed.

ABSTRACT. 143 undergraduates studied an expository text for four minutes in either a control condition, or while getting interrupted repeatedly to complete simple secondary tasks. Interruptions increased self-reported difficulty and decreased comprehension. The effect was smaller when participants were permitted to delay their response to the secondary tasks. Implications for a theory of reading goal management are discussed.

ABSTRACT. Eighty-six adult literacy learners shared their utility value for writing improvement in grammar and spelling, word processing, and planning, drafting, and revising. Findings indicated that they possessed a high utility value for improving writing in all areas and had a variety of education, work, and personal motivations. Their age and educational attainment were related to utility value for grammar and spelling improvement, and age was also associated to utility value for improving word processing skills.

ABSTRACT. This study uses Natural Language Processing (NLP) to analyze changes made by the College Board to their AP African American Studies curriculum. It finds the updated version may have been influenced by political pressure against "divisive" concepts, based on the removal of topics, themes, names, and sources relevant to African American Studies. This study demonstrates NLP's utility for curriculum inquiry, and highlights future work needed on student learning in response to curriculum text.

ABSTRACT. Complex language in ballot measures reduces engagement in the civic process. Even supplementary information about ballot measures can be too complex for the average reader. Participants (n = 192) read ballot measures and supplemental summaries in their original form or in one of three simplified versions (surface, Plain Language, discourse-informed). Results showed that simplification improved readers’ perceived comprehension and comprehension, but had minimal effect on whether they chose to cast a vote or opt out.

ABSTRACT. The current study identified writing profiles of source-based essays (n = 2,643) to contribute to the development of a writing analytics tool. Linguistic features of the essays were extracted using Natural Language Processing tools, emphasizing lexical, syntactic, and discourse features. Principal Component Analysis revealed nine linguistic components explaining 42% of variance in writing quality. Components related to lexical diversity and noun phrase complexity and diversity were particularly influential on source-based writing quality.