ABSTRACT. In this study, we examine whether Kindergarten, Grade 1 and Grade 2 children (N=142) can make elaborative and predictive inferences while listening to fictional stories. Inferential skills were assessed through a drawing decision task. Our result showed that most children were able to produce both types of inferences. But, more precisely Grade 1 and Grade 2 spontaneously focused more on predictive inferences whereas kindergarten were more accurate on elaborative ones.

ABSTRACT. This project examined how positive versus negative emotions influenced integration and memory for text. This study utilized an emotion induction procedure followed by an experimental reading task from O’Brien and Cook’s inconsistency paradigm (2014). The results replicated traditional inconsistency effects. In addition, positive emotions had a delayed facilitative effect on integration during reading. After reading, positively induced participants were more likely to comment on the inconsistency during immediate recall, but less likely during delayed recall.

ABSTRACT. The ability to assess another person’s understanding is necessary for successful collaboration. In online collaborations we lack access to the traditional cues, such as speech delivery and gesture, we use to inform our perception other’s knowledge base. As a first step toward exploring the process of knowledge estimation in online shared workspaces, in this study we examined how typing speed and typing disfluencies shape viewers’ perception of the typist.

ABSTRACT. Previous research has shown that accompanying text to artworks (e.g., titles) increases the preference for artworks. We further examined the influence of the presence and cohesion of artwork descriptions on aesthetic responses. We found that accompanying text increased appreciation and that the effect of cohesion depended on the viewer’s interest in art. Less interested viewers were more affected by cohesion than more interested viewers.

ABSTRACT. This study investigated L2 readers’ monitoring of local and global coherence along the protagonist, causal, and intentional dimensions of narratives. Eighteen Japanese university students read English narratives containing context and target sentences separated by one (the local condition) or four filler sentences (the global condition) with their eye movements recorded. The results showed that readers constantly monitored protagonist coherence; however, they had increased difficulty monitoring causal and intentional coherence.

ABSTRACT. Purpose: To examine the effect of disfluencies on perceptions of speech quality. Method: Audio samples were gathered from 20 bilingual speakers and rated by unsophisticated listeners using subjective fluency and quality measures. Results: Ratings were highest for speakers with a higher number of disfluencies. Analyses indicate a significant difference between the low and average rated groups. Conclusions: Higher disfluencies may contribute to better listener recall and positive listener perceptions of speaker effort for bilingual speakers.

ABSTRACT. This experiment tested whether performing a pre-search task involving either the spelling or the meaning of words modified the way 11 years-old children subsequently scanned a text to find a single-word answer to a question. Compared to adults performing the same searches, eye movement recordings revealed that the induction tasks had a stronger influence on young adolescents. In particular, the semantic induction task may facilitate question-answering by adolescents when the text contains semantically-relevant distracting information.

ABSTRACT. In this study, students participated in online debates as required by an online course. Each posting was requested to tag into one of four categories: (1) an argument that supports or opposes the main claim, (2) a challenge, (3) an explanation, or (4) supporting evidence. Epistemic network analyses were used to identify reasoning patterns and their dynamic change over a series of online debates. Implications are discussed in terms of collaborative on learning and discussion.

ABSTRACT. Texts vary. Accordingly, do different texts encourage certain types of online comprehension processing? This presentation illuminates how science texts with varying levels of cohesion may contribute to the online comprehension processing of seventh grade readers during a think-aloud task. Our analyses illustrate how students’ inferential processing differed in science texts with varying degrees of deep and referential cohesion. Implications are drawn about the effects of text cohesion for online inference generation in adolescents.

ABSTRACT. Within the RI-Val model of reading comprehension, the coherence threshold marks the point at which the reader has deemed comprehension sufficient to move on in a text. Previous research has demonstrated that the readers’ coherence threshold can be manipulated by increasing task-demands (Williams et al., 2018) or including text-based disruption in coherence (Sonia & O’Brien, in prep). The goal of the current research was to investigate the resetting of the coherence threshold to baseline.

ABSTRACT. Researchers in experimental aesthetics are interested in people interpret and engage in visual discourse, such as artworks. Unsurprisingly, people’s willingness to cognitively engage with art may be a key factor, but there are not yet measures that capture this construct. The current research proposes the Aesthetics Processing Preference Scale (APPS) to fill this gap. In two studies, the APPS was found to be a reliable and valid measure of people’s willingness to engage with art.

ABSTRACT. Example-based instruction with text involves introducing readers to domain principles (e.g., principles of natural selection), followed by several examples that illustrate those principles. It is unclear how bridging inferences contribute to example-based learning. Participants did think-aloud while reading four short texts: one about principles and three about examples of those principles. The main finding was that bridging inferences to the principles were predictive of learning, whereas bridging inferences to the examples typically were not.

ABSTRACT. Effective computerized reading comprehension strategy training requires an ability to provided automated an accurate to students as they answer open-ended comprehension question. This study explored different approaches to machine scoring as part of a larger research and development project. Two units involving 4 open-ended questions were used for initial testing. A comparison of simple frequency and deep learning techniques suggest that the latter have more potential to provide accurate feedback on response correctness.

ABSTRACT. This study examined the nature of individuals’ thoughts during reading. We examined whether self-reports of thought characteristics (i.e., mind wandering, visual, verbal, valence) varied across time and task. In two sessions, participants (n=58) responded to thought probes across focused meditation and narrative reading tasks. Results showed that participants’ thought patterns were stable across sessions. Further, reports of mind wandering, visual imagery, and valence varied based on task, whereas verbal thinking remained stable across tasks.

ABSTRACT. We computationally emulated a link between symbolic and emotional representations of words using computational models and predictive models. We studied dimensional and discrete emotions using two different predictive models: linear regressions and neural networks. More than 13000 words were used to train the models and then they were tested in more than 4000 words. While important differences were observed between linear regressions and neural networks in dimensional emotions, no differences were observed in discrete emotions.

ABSTRACT. We examined hierarchical relations and dynamic/differential relations of language and cognitive component skills to writing – whether the relations of component skills to writing are direct and mediated, and whether the relations vary as a function of dimensions of written text. Structural equation model results from 350 second graders showed that not all component skills were directly related, and total effects of component skills (e.g., working memory, vocabulary, perspective taking, monitoring) varied largely.

ABSTRACT. High school students (N=389) read multiple texts on the topic of American football in a scenario-based learning environment. Learning was evaluated with a pretest-learning-posttest design. Comprehension questions were asked throughout the learning environment. Topical knowledge and reading ability were also evaluated. Comprehension was the strongest predictor of learning. Those who spent longer time learning learned better. Reading ability and knowledge positively predicted learning. High reading ability compensated for low knowledge in learning.

ABSTRACT. A methodological framework for collecting metacognitive information from students with significant cognitive disabilities is applied in the evaluation of the Georgia Alternate Assessment 2.0. This methodology blends concurrent and retrospective cognitive interview practices in an evaluation of verbal and nonverbal responses from students and educators. Educator-mediated communication is prioritized, and these results, in conjunction with operational results from over 13,000 K-12 students, validate the intended assessment design and the viability of this blended methodology.

ABSTRACT. This study examined whether inferencing mediated the relations between language-based component skills and reading comprehension controlling for other lower-level skills. Word reading fluency and vocabulary knowledge were predictive of the sentence-level comprehension. Inferencing mediated the relation between vocabulary and the passage-level comprehension. Component skills varied as a function of comprehension measure administered. These findings suggest a need to administer multiple measures of comprehension to understand the underlying component processes involved in adults’ reading comprehension skills.

ABSTRACT. This study aimed to examine the role of executive control in the integrative reading processes as well as the learning outcomes in illustrated science text reading. Twenty-eight 4th/5th grade elementary students read two illustrated texts while their eye movements were recorded with an eye tracker. Findings were a) integrative transition is associated with their transfer outcomes, b) attention shifting is associated with learning outcomes, and c) no working memory capacity is associated with integrative transition.

ABSTRACT. Morphological awareness is a significant factor in reading comprehension; and struggling readers may use it as a compensatory reading strategy. A Morphological Awareness Battery (MAB) was developed and tested to determine whether it could predict reading comprehension of fourth and fifth grade students with significant reading deficits. The MAB significantly predicted between 10% and 53% of the variation in passage reading comprehension and 35% of the variation in passage reading fluency. Intervention implications are discussed.

ABSTRACT. Task-unrelated thought (TUT) occurs frequently in our daily lives, but we know little about how it influences our communication. The current study explores how frequently our minds wander during text-based computer-mediated communication and how group membership impacts this rate. Participants reported TUT once every two minutes on average, but this rate was lower when participants perceived talking with an ingroup member. More frequent TUT was also related to more frequent topic shifting across all conditions.

ABSTRACT. We examine changes in discourse patterns over time in an argumentation intervention for adolescents. The intervention was designed to promote text-based classroom discussion via highly-supportive instructional materials including queries to support comprehension of argument texts and analysis of genre features. Over nine weeks, patterns (rather than frequencies) of teacher moves revealed that only in intervention classes, teachers used fewer clusters of discourse moves that suppress dialogic reasoning. Student talk to build reasoning across participants increased.

ABSTRACT. This poster describes a unique web-based program with embedded AI to individualize reading comprehension instruction that helps students learn STEM content. The underlying learning pedagogy is based on the CI model of comprehension. The project seeks to understand student behaviors to build algorithms that predict and deliver only the instruction each student needs when they need it as they learn from text to build their expertise. Relevant students’ behavior patterns is discussed.

ABSTRACT. The purpose of this study was to examine how learning words with matching and mismatching representational gesture affects subsequent comprehension of these newly-learned words within read sentential contexts. Results did not support our hypothesis that pseudowords learned via text and co-occurring semantically-congruent representational gestures would enhance subsequent identification and processing when reading them in sentential contexts. Implications are discussed.

ABSTRACT. Critique and evaluation of scientific evidence and alternative explanations about a phenomenon are essential for students’ participation in many scientific practices and can lead to deeper scientific learning. The present study found that instructional scaffolds that afforded students’ greater conceptual agency related to higher levels of scientific evaluation, increased plausibility shifts toward the scientific, and deeper understanding of socio-scientific issues, including causes of climate change, availability of water resources, and origins of the universe.

ABSTRACT. In this project, we evaluated the quality of written arguments through analyzing examples in test takers’ essays in an English language assessment. Altogether we identified 168 examples used to support arguments in 99 essays. Raters were able to recognize various characteristics of the examples in a relatively consistent manner. The results indicated that the number of examples and clarity were significant predictors of the essay quality. Here, we present our analysis approach, results, and implications.

ABSTRACT. The Early Language Comprehension Individualized Instruction (ELCII) application uses video-based inferential questions and scaffolding to teach inferencing to kindergarteners. In this study, we investigated the efficacy of ELCII's scaffolding to facilitate inferencing performance, while also accounting for individual differences in language comprehension and executive function. Results show that scaffolding facilitated inferencing for all students, but provided slightly more benefit for students with lower executive function and language comprehension skills.

ABSTRACT. This study extracted content and structural features to predict human annotations for claims and non-claims in argumentative essays. The evaluation of classification models indicated Gaussian Naive Bayes classifier yielded the most balanced identifications of claims and non-claims. We used the model to make predictions in a validation corpus that included human ratings of writing quality. The number of claims, the percentage of non-claims, and the average position of non-claims were significant indicators of essay quality.

ABSTRACT. We assessed progress-monitoring data from TeLCI, a Technology-Based Early Language Comprehension Intervention, to identify the impact of diverse and marginalized backgrounds due to race, socio-economic status, and home language. Students from Grade 1 and 2 (n = 62) with language comprehension difficulties completed the 24-module program to build inference-making skills without relying on decoding. Two-way ANOVAs and non-parametric tests reveal an achievement gap for English language learners and students of low socio-economic status.

ABSTRACT. New U.S. science standards require the negotiation of teaching principles and practices using an anchoring task across a unit. The current studies explored classroom discourse to describe how teachers negotiated these elements. Teachers primarily discussed the anchoring task, with few mentions of the principles and practices. Unsurprisingly, students thought the goal of the unit was the anchoring task. We discuss the importance of framing the anchoring task as support for learning the principles and practices.

ABSTRACT. This project examined whether exposure to accurate and inaccurate information in fiction influences readers’ confidence in judging the validity of related claims. In Experiment 1, participants made more judgment errors after reading inaccurate versus accurate information, and were less confident in their incorrect than correct judgments. In Experiment 2 we tested whether confidence could be leveraged to reduce incorrect judgments. Allowing people to withhold judgments did not eliminate the consequences of exposure to inaccurate information.

ABSTRACT. Students often struggle to make sense of literary works. We used a think-aloud design to examine the effect of four reading instructions (rules of notice, rules of signification, combined, control) on students’ processing of a literary short story. Protocols were assessed for evidence of expert-like reading behaviors such as attending to literary features and generating interpretive inferences. We also examine how differences in reading behaviors related to students’ appreciation and enjoyment of the work.

ABSTRACT. Many sources that report about scientific issues are highly partisan and differ in their treatment of scientific topics. The present study examines how different sources (liberal vs. conservative vs. scientific) and tentative language (hedged vs. certain) influence liberals’ and conservatives’ belief in scientific information and trust in sources. Results indicate that liberals believed and trusted scientific sources more than partisan sources, whereas conservatives believed and trusted conservative sources as much as they did scientific sources.

ABSTRACT. The Knowledge Revision Components Framework (KReC) describes how a refutation text facilitates revision of misconceptions. However, readers frequently engage with multiple documents, but we lack understanding of knowledge revision in multiple-document contexts. Thus, we propose a new framework, KReC-Multiple Documents, to predict how factors such as source credibility and multiple-document integration influence revision with multiple refutation texts. Results indicate that high-credibility sources and greater multiple-document integration facilitate better knowledge revision, illustrating initial viability of KReC-MD.

ABSTRACT. The current research examined if readers detect and attempt to resolve coherence breaks within scientific texts. Participants read explanations that did or did not contain an inconsistency. Eye-tracking data showed readers spent more time processing inconsistent sentences than consistent sentences. Participants did not show awareness of the inconsistencies after reading, and infrequently drew inferences to repair the issue. Readers appeared to experience a coherence break, but frequently dismissed or were unable to resolve the issue.

ABSTRACT. Belief in scientific information may be influenced by the source of the information, the language used, and readers’ epistemic beliefs (EBs). We examined the influence of tentative language (hedged vs. certain), source (liberal vs. conservative vs. scientific), and three dimensions of EBs (Faith in Intuition, Need for Evidence, and Truth is Political) on belief in climate-change information. We found interactions among epistemic beliefs and source and language on belief in scientific information. Implications are discussed.

ABSTRACT. Validation is an integral part of text comprehension. We used reading times and plausibility judgments to investigate combined effects of source credibility and plausibility on validation. Participants read stories with a high- vs. –low-credible person making knowledge-consistent, implausible, or knowledge-inconsistent assertions. Interactions of source credibility and plausibility were found for plausibility judgments and reading times, indicating that source credibility affects validation but that the pattern of effects depends on the degree of implausibility.

ABSTRACT. The current study explored the utility of a computerized program called the Graphical Interface of Knowledge Structure (GIKS) that generates and compares a network from a student essay and a master text. We compared different structures of master texts on the same content and also compared GIKS-identified nodes and links to those scored by human experts. We found that GIKS was able to improve node identification from essays by using regular expressions.

ABSTRACT. The Predict-Observe-Explain (POE) learning cycle improves understanding of the connection between empirical results and theoretical concepts when students engage in hands-on experimentation. This study explored whether training students to use a POE strategy when learning from social science texts that describe theories and experimental results might be more beneficial than an explanation strategy. The study found that students trained to use an explanation strategy displayed better comprehension on a new set of topics.

ABSTRACT. In previous research, participants engaged in retrieval practice, or simply reread, texts containing seductive (interesting but irrelevant) details. Participants retained more information after retrieval practice, but only for seductive details, not important information. Here, we conducted the same comparison after removing seductive details from the texts. Participants retained significantly more important information after retrieval practice compared to rereading. Seductive details seem to affect processing both during reading and in retrieval practice engaged after reading.

ABSTRACT. Seductive details are highly interesting, but irrelevant, elements added to learning materials. The current study examined how the inclusion of these details influences learners’ ability to form inferences between texts. Participants were asked to read two texts about El Niño and to judge the validity of inferences from across texts. Results show that seductive details lead to decreased performance, though learner characteristics provided some general benefits on this task.

ABSTRACT. 1374 undergraduates took a shorter version of the verbal-reasoning section of LSAT test and a justification of scientific knowing questionnaire. A principal component analysis yielded three dimensions: Personal Justification (JP), Justification by Authority (JA), and Justification by Multiple Sources (JMS). Whereas students who relied highly on JMS performed better on verbal-reasoning task than their less-relying counterparts, JP had the opposite effect. Implications of the results and validation of the justification of knowing questionnaire are discussed.

ABSTRACT. This paper examined the learning effect of word appeared in the 1st and 2nd graders’ Chinese language textbooks. The words were selected by frequency in the textbooks, frequency in corpus and by their categories. It was found that the words shown in the textbooks were not challenging to the urban school students. The results and following study will be discussed in this report.

ABSTRACT. The main purpose of the present study was to investigate the processing changes in constructing three levels of mental representations while reading fully versus partially overlapping narrative passages by Korean EFL learners. This study employed the Resource Allocation Approach (Lorch & Myers, 1990) as a theoretical framework. Results showed that the full repetition significantly facilitated the construction of the surface form and the textbase, whereas the partial repetition promoted the development of the situation model.

ABSTRACT. Eye movements are a valuable source of lexical processing information. With eye-tracking technology, word reading data can be standardized and used as lexical processing benchmarks for text analysis. This study tests this application by extracting monolingual and bilingual eye-behavior information from an eye-behavior corpus (GECO; Cop, Dirix, Drieghe, & Duyck, 2017) to predict reading comprehension of first (L1) and second language (L2) readers. The eye-behavior norms were successful predictors of both L1 and L2 reading.

ABSTRACT. This study investigated how the use of cohesive devices predicts writing fluency for second language (L2) undergraduate students (N = 99). Linear mixed effects models were built to predict writing fluency using cohesion indices. Results showed that the use of semantic overlap between adjacent sentences negatively predicted writing fluency. The use of more unattended demonstratives predicted higher production rate but greater revisions, whereas the use of more attended demonstratives predicted fewer revisions.

ABSTRACT. This study investigated the effect of text cohesion on second-language reading comprehension. Twenty Japanese undergraduates read both low- and high-cohesion texts in English and performed free written recall. The results showed that low-proficiency readers recalled more information from high-cohesion texts than from those with low cohesion. However, highly proficient readers had the opposite tendency, indicating that the benefits of high cohesion are restricted to relatively poor readers while proficient readers are assisted by low-cohesion texts.

ABSTRACT. Various scholars have previously examined right-wing rhetoric and mass media in the United States and Europe, particularly in regard to specific issues such as anti-immigration stances. Similarly, this paper uses critical discourse-analytic methods to study a corpus of 27 anti-immigration news and opinion articles from right-wing websites. The articles were published at three particular junctures in the administrations of President Obama, and within the first three years of the Trump administration.

ABSTRACT. This study developed a literariness analysis framework and then invited 22 experts to assess the importance of the aspects, criteria, and indicators of this framework. Data were analyzed using the analytic hierarchy process. The results show that the five most important indicators were innovation, emotional genuineness, attitude, life experience, and topic. The importance for text literariness of the indicators in this framework differed, with the global weight being highest for innovation and lowest for substitution.

ABSTRACT. In this study, we examined the performance of young children making inferences in non-reading contexts across two different genres. First- and second-grade students participated in an intervention that involved comprehension activities using fiction and non-fiction videos in a technology-based environment. There was no genre difference in students’ inferencing performance after scaffolded feedback. This result suggests that knowledge demands in the non-fiction genre were likely minimized as a function of video-based scaffolding.

ABSTRACT. Expository texts are considered more difficult to comprehend than narrative texts. This study’s purpose was to examine two potential reasons for this genre difference: causality and intentionality. Based on results of mediation analyses, causal content did not mediate the effect of genre on comprehension accuracy perhaps. However, intentional content explained some of the effect of genre on item accuracy. One interpretation is that protagonist goals in intentional content guide readers to better understand narrative texts.

ABSTRACT. This study explores the impact of editing and shot-scale on film comprehension. Participants (n = 120) viewed a film depicting a man and woman performing a modern dance. Two versions employed editing to focus on either the male or female dancer; a third version was from a single objective camera position. After viewing, participants were asked to recall what they saw. These recalls were analyzed to examine how the editing choices influenced viewers’ sense of narrativity and character.

ABSTRACT. Previous think-aloud research with children found differences in online processing between expository and narrative texts (e.g., Karlson et al., 2018). We sought to extend these findings to adult readers. We compared responses generated by college students during think-aloud tasks. Our adult readers showed similar differences between text genre as found by previous studies with children. With expository texts, readers produced more associations, metacognitions, and text connections. With narratives, readers produced more valid elaborations and predictions.

ABSTRACT. This study argues that new metrics for assessing debate performance, including floor control and idea resonance, provide a stronger framework for gauging success than the traditional markers. Using a new technique for thematic transcription of a debate between Supervisor Harvey Milk and Senator John Briggs over a proposed law which would have removed gay teachers, it rethinks debate success and posits how this analysis be used to think through debate performance today.

ABSTRACT. Natural language quantifiers — words such as few, a few, and many — play a powerful role in influencing readers’ focus. The polarity of a quantifier, positive (a few) or negative (few), largely determines focus effects. In a series of experiments, we demonstrate that a naturalistic story context can influence the interpretation of positive quantifiers. Although positive quantifiers almost always lead to focus on the reference set, this interacts with the pragmatics of the discourse.

ABSTRACT. Adding cues such as connectives and word overlap to text increases cohesion, and in turn, text ease. In this study, Russian students (ages 11-12; n = 65) read either a low cohesion (i.e., original version) or high cohesion (i.e., modified) informational text. Replicating prior studies conducted in English, the students who read the high cohesion form of the Russian text recalled more propositions from the text than those who had read the low cohesion form.

ABSTRACT. In this study, we analyze comprehension and metacomprehension differences between expository (EX) and informative narrative (IN) texts when answering different questions following Kintsch’s comprehension model (textbase/situation model). 100 participants read different expository texts and answered multiple-choice questions. Expository texts were presented in two formats (EX/IN). A mixed-effects model showed no interaction between structure of text and type of text, but a statistically significant effect of text structure. Results are discussed according to previous research.