ABSTRACT. Reading skill is foundational to academic and occupational success and therefore understanding the factors that support successful reading development is critical. The current study investigated whether neural specialization for phonological and semantic processing at 5-to-6.5 years old predicts growth in reading skills at 7-to-8.5 years old. Results from this preregistered study provide important preliminary evidence in favor of the role of early phonological neural specialization in the development of word reading skills.

ABSTRACT. Self-explanation is designed to increase coherence by encouraging students to activate prior knowledge, generate inferences, and make casual connections (McNamara, 2004). The current study used natural language processing to examine how readers’ responses differ when instructed to self-explain or think aloud. Self-explanations were found to contain more cohesion, semantic overlap, and causal, active, and positive emotion words than think-alouds. The results provide evidence that instructional differences significantly predicted linguistic differences reader’s responses to texts.

ABSTRACT. Morphological accuracy, complexity, and awareness are often considered important benchmarks in language acquisition and performance. Though morphology is underexplored in natural language processing, automatic measurement of morphological complexity in English can lend insights into various aspects of text and discourse processing. This study introduces a tool to automatically process morphological complexity in texts. Spoken and written English-learner corpora were analyzed using the tool to explore the relationship between morphological complexity and discourse types.

ABSTRACT. This study compares the performance of five different traditional and new readability formulas in the task of classifying simple Wikipedia and authentic Wikipedia articles (N = 4,000). Results indicated that a new formula, the Crowdsourced Algorithm of Reading Comprehension (CAREC) performed the best. The traditional readability formula, Flesch-Kincaid Grade Level, also showed reliable performance. The results suggest the linguistic features used in newer readability formulas are capable of reliably representing the difficulty of a text.

ABSTRACT. In this study, we proposed two approaches to summary analysis (model-based and text-based) along three dimensions: surface, structure, and semantic. We investigated the power of the two approaches to assess changes in students' summaries. Results demonstrated the theoretical overlap of model-based and the text-based approaches and the potential for a more nuanced account of how students understand text.

ABSTRACT. Based on serial reproduction experiments, Frederic Bartlett (1932) suggested that the stereotypical form of narratives consists in rationalization, meaning causal connections. We conducted the largest retelling experiment to date (18,738 retellings) that suggest that affects, and especially the precise preservation of the story’s degree of happiness and sadness, survive retelling with few changes, while many aspects related to fact preservation, coherence, and rationalization of the story deteriorate. We speculate about the function of narrative communication.

ABSTRACT. Recent research on digital literacy has examined how multimodal projects can enhance multilingual students’ academic writing (Bloch, 2018; Spina-Caza & Booth, 2011). However, little documentation exists on how students’ informal use of digital multimodality, as in social media, may affect their academic writing. This presentation will give an overview of digital multimodality as informally used by second language writers and share how rhetorical abilities L2 writers already have can be harnessed to bolster academic writing.

ABSTRACT. We explored whether self-explanation strategies influence the “screen inferiority effect” (superior comprehension after reading on paper compared to digital media). Participants read a text on paper or on a computer monitor, and received instructions to self-explain or to read for comprehension. On a subsequent comprehension test we observed a screen inferiority effect, but only in the self-explanation condition. Analyses of processing times, metacognitive judgments, and constructed responses are presented to help explain these results.

ABSTRACT. Museums have been exploring the potential of augmented reality (AR) as a means to promote science engagement. This proposal reports on the design and initial test of an AR exhibit at an active paleontological dig site. Results from analysis of participants’ discourse in response to the experience show that AR increased visitor interest and positive emotions around science content. Significant learning gains and decreases in science misconceptions also occurred for participants.

ABSTRACT. This study’s purposes are to examine whether reading medium and interest, both individual and situational, interact to predict performance on a reading assessment and whether medium affects situational interest. College students (N = 206) reported their individual interest, were randomly assigned to read a textbook excerpt from paper or screen, and then reported their situational interest. Situational interest was more predictive of performance from screens than from paper. Medium did not influence situational interest.

ABSTRACT. As politics takes over more and more of the public discourse, what information influences citizen’s voting choices has become increasingly relevant. Comedic news shows are one such source of knowledge. This experiment endeavored to determine whether humor in these shows enhanced or inhibited memory of their factual contents. Participants remembered more when shown a segment with jokes than the same segment without them, with liberals out-learning this liberally-leaning content compared to their conservative peers.

ABSTRACT. People enjoy the affective response engendered by filmmakers through narrative. In two experiments, we tested the role of film audio and working memory on a predictive inference important for narrative suspense. Participants watched three minutes of Touch of Evil. We manipulated knowledge of a time-bomb when the scene starts. Audio increased the likelihood of generating a bomb related inference (Experiment 1). Participants higher in working memory were more likely to generate bomb inferences (Experiment 2).

ABSTRACT. The Early Language Comprehension Individualized Instruction (ELCII) program uses video-based inferential questions and scaffolding to train inferencing in kindergarten. We predicted that posing questions during comprehension of videos (online condition) would lead to better performance than posing questions afterwards (offline condition). Moreover, we predicted that higher executive function (EF) would facilitate greater growth in inferencing skill. Results revealed that students made greater gains in inferencing in the online condition, with high-EF students demonstrating greater gains.

ABSTRACT. It is known that eye movements during reading reflect various reading processes as well as reader skill and attentiveness, but there is little work relating eye movements to reading comprehension outcomes. This work represents a novel step by showing that deep comprehension assessed by open-ended self-explanations during reading (r=0.32, p<0.001) can be predicted from eye movements in a person-independent manner. Our results have implications for theories of reading and for the design of real-time interventions.

ABSTRACT. This study is concerned with the question how endogenous eye movement dynamics change as they become contingent on external (linguistic) information. It is hypothesized that external information lead to increased sequential order of eye movement measures, compared to conditions that contain little or no information. To test this hypothesis, eye movements of 26 German native speakers were recorded during reading-unrelated and reading-related tasks. To analyze the data, we used recurrence quantification analysis (RQA), which quantifies the degree of temporal structure in time series. Recurrence measures of eye movements convincingly distinguish between conditions. Findings suggest that qualitatively different tasks can be measured on a continuum of temporal structure and provide new perspectives for further studies investigating natural reading as complex, dynamical process.

ABSTRACT. This study investigated how causal and semantic relatedness between sentences affects second language discourse processing as reflected by eye movements. Japanese learners of English read two-sentence texts varying in causal and semantic relatedness and their eye movements were recorded. Linear mixed-effects models of eye movement measures revealed that causal relatedness has a robust impact on both fixation durations and lookback frequency, whereas the effects of semantic relatedness are modulated by causal relatedness and reading skill.

ABSTRACT. This eye-tracking study examined how differences in sources’ trustworthiness affect readers’ attentional processing when confronted with a scientific conflict. 144 participants were presented with two conflicting scientific claims from two sources. Results show that differences in trustworthiness between the two sources led to increased attention to source information compared to when both sources were of high trustworthiness or of low trustworthiness, which we interpret as an indication for conflict resolution via sourcing.

ABSTRACT. We present an integrated model of individual growth (multilevel SEM) to examine 10,701 reading times from 20 to 24 sentences each in four texts read by 123 college students. We evaluate the extent to which reading times indicate a single cognitive process, common across texts, versus distinct trends which suggest texts invoke different, distinctive cognitive processes. Findings suggest interesting commonalities as well as distinct features of sentence, text, and person-level features.

ABSTRACT. We examined the limitations of the spatial gradient effect, or decrease in availability of objects/locations as a function of distance traveled. Across three experiments we used naming time probes to measure availability of an initial spatial location after varying the distance a protagonist traveled. The findings will be discussed in terms of the limit they place on memory-based explanations, and the need for dimensional information that is a part of situation-based explanations of comprehension.

ABSTRACT. Though the roles of working memory (WM) and prior knowledge (PK) in reading comprehension have been studied extensively, their effects are rarely studied concurrently. Much of this work has struggled to adequately assess WM or has used insufficient measures of comprehension. The present study simultaneously tested the impact of WM, vocabulary, and domain-specific PK on reading comprehension. Only domain-specific PK predicted unique variance in reading comprehension, emphasizing the importance of PK for building understanding.

ABSTRACT. This study revisits the effects of text complexity and individual differences on comprehension in the context of the Globally-Integrated Scenario-based Assessment (GISA). High school students (n = 511) completed prior knowledge and foundational reading skill assessments followed by either a high or low cohesion version of a scenario-based comprehension assessment. Preliminary analyses indicate robust effects of prior knowledge, little effects of foundational reading skills, and no overall effects of the cohesion manipulations on comprehension performance.

ABSTRACT. We tested whether readers' attention to source information depends on their prior knowledge. 102 undergraduates read true, false, or uncertain statements attributed to either competent or less competent sources. Participants rated the statements as true or false. Statements attributed to competent sources were more likely to be rated as true, but the effect was much larger when the statement was uncertain than when it was either true or false. Implications for validation processes are discussed.

ABSTRACT. After reporting initial beliefs, subjects read a belief consistent or inconsistent text about gun control effectiveness. Subjects verified initial beliefs about gun control that were either accurate, the opposite of their initial belief (misinformation), or did not verify. 80% of misinformation subjects thusfar verified an incorrect belief as their own. Subjects significantly change beliefs about gun control after reading a belief inconsistent text compared to a belief consistent text. There was not an overall influence of the verification condition on post-reading beliefs.

ABSTRACT. We examined whether readers’ revised knowledge as a result of reading refutation texts can transfer to texts that were designed to (a) re-activate or (b) re-activate and background misconceptions that were addressed in the refutation texts. Reading time results show that reading a refutation text facilitates transfer of readers’ revised knowledge to the next text (Experiment 1). However, as the distance between refutation and transfer texts increases, transfer of revised knowledge is reduced (Experiment 2).

ABSTRACT. Readers’ preferences for what happens in a text affect both comprehension and memory of story outcomes. In general, people take longer to read stories in which outcomes are inconsistent with what they wish would happen. But, demonstrated in this project, negative preferences exhibit specific and important effects. After establishing preferences for character failures, participants take longer to read outcomes, and have poorer memories for those outcomes. Readers preferences are routine, exemplifying situation model contents.

ABSTRACT. Discipline-specific literacy is necessary for all learners to develop a sense of competency and interest in those fields. To develop such proficiency, learners must also have access to appropriate linguistic resources. Considering DHH learners, we have identified a lack of ASL resources that can support them in communicating in ways emphasized for science classrooms. Further work with DHH learners indicates that this creates a sincere barrier to developing affinity and interest in the STEM fields.

ABSTRACT. This presentation describes a process for developing a literacy assessment system tailored to the needs of d/hh students in K-12 settings. Over three years, the researcher facilitated a process with teachers that included introducing, evaluating, selecting and using assessments to create individual and class profiles that guide instruction. Reflections on the construction of an assessment system as a capacity-building activity, and results from two years of data collection, analysis and instructional decision-making will be shared.

ABSTRACT. The narrative production abilities of 32 deaf children between third- to sixth-grade from two American Sign Language (ASL) bilingual schools will be presented. The participants were exposed to two narrative conditions of wordless picture books, one was produced in ASL and the other was conducted in Simultaneous Communication. Participants then retold the narrative. The findings will discuss the quality of the narrative structure and whether they followed the same communication modality of the initial exposure.

ABSTRACT. Standardized testing values and purposes continue to be debated (Boaler, 2003; Qi & Mitchell, 2011). The purpose of this paper is to show the importance of the role of the principal in understanding the impact of standardized testing on deaf students. A case study of a school principal and how he/she reviewed the text and questions being asked on a standardized test to demonstrate the need to improve testing for deaf students will be presented.

ABSTRACT. Academic English skills of bilingual d/hh middle and high school students is a profoundly understudied area. Few studies have examined this skill within this population, and have found a relationship between American Sign Language proficiency and both reading and writing in academic English (Scott & Hoffmeister, 2017, 2018). This presentation will share results from a study that expands on these findings by exploring the role of early and late language exposure on academic English skills.

ABSTRACT. A large portion of adults struggle to read at a basic level. While foundational component reading skills (e.g., decoding, vocabulary, morphology, sentence processing) are known to account for a large portion of variance in reading comprehension, this study used the Reading Systems Framework to explore potential interactions between component reading skills. Results suggest that word-level processes interact with lexical knowledge in predicting comprehension among struggling adult readers.

ABSTRACT. High school students with significant reading deficits must read to learn course content. A pilot study of Readable English, a phonics intervention providing embedded interactive orthography to scaffold online grade level content, significantly increased both reading accuracy and reading comprehension compared to control group (N=24, mean age = 16.5, >1.5 years below grade). Findings suggest this new learning intervention technology may particularly benefit students struggling to read and pronounce English at the word level.

ABSTRACT. To be valid, models of reading disability should predict important real-world criteria. An important real-word criterion for students with reading disability is whether they would profit from text-to-speech. Data collection is currently underway ending after Spring 2020. Then, we will compare alternative models of reading disability by their capacity to predict reading individual differences in reading comprehension performance using TTS, controlling for unassisted reading comprehension.

ABSTRACT. Computational semantic space models have now been applied to sentences, but it is unclear whether they capture how the human brain represents sentences. Using fMRI we scanned adult readers reading expository texts and compared their brain responses to 3 semantic space vectors that modeled sentences either as combinations of words or as single units. We observe that computational semantic representations that are specifically designed to capture sentence content share information content with brain responses.

ABSTRACT. A large percentage of the US adult population struggles with basic reading skills, but there are few valid assessments designed for them, making it difficult to measure learning outcomes or improve instruction. To remedy this, we are developing digital assessments appropriate for adults with below-basic literacy skills. Such assessments will not only help to determine an adult reader's strengths and weaknesses but also inform instruction and improve program and institutional accountability.

ABSTRACT. We examined individual differences that contribute to source-based essay writing in an adult literacy population. Participants (n=143) wrote a source-based essay and completed a battery of individual difference measures related to literacy. Results indicated that the quality of source-based essays was differentially predicted by individual differences related to language knowledge, comprehension ability, and persuasive writing skill. These results suggest that source-based essay writing is a complex task that relies on a host of developed skills.

ABSTRACT. The persuasive writing of N=65 college developmental education/ remedial students was compared to that of N=72 typically-performing undergraduates and N=112 Masters students. Twelve variables covering writing quality, vocabulary usage and linguistic aspects of writing were analyzed. There were virtually no statistically significant differences between native and non-native speakers. A series of cluster analyses suggested that the data converged into two clusters. Overlap between cluster and educational placement varied in unexpected ways.

ABSTRACT. MOCCA-College is a new version of the MOCCA cognitive diagnostic reading comprehension assessment, which differentiates between 2 subgroups of struggling comprehenders (i.e., paraphrasers and elaborators). Preliminary results from post-secondary students (N=1704) indicated strong internal reliability. Results also indicated weak-to-moderate positive correlations with other reading assessments. Qualitative and qualitative item analyses using IRT, Coh-Metrix, and human coding provided insight to item difficulty and discrimination—guiding item and form revisions for subsequent years of the 3-year study.

ABSTRACT. An alarming number of first year college students are underprepared, and in particular with respect to being ready to read in college. This study explored the factors that predict success on literacy task and early academic success. Specifically, this study explores the extent that foundational skills associated with reading, strategy use, and motivation for reading account for variance in academic reading and success over and above traditionally measures (i.e., ACT, SAT, GPA).

ABSTRACT. Although frequently used with expository texts, illustrations can lead to illusions of understanding. When students studied geoscience texts without sketching, both comprehension and monitoring were poor if only some topics in a set were illustrated. However, when students were prompted to generate a sketch while reading, both comprehension and monitoring were improved by sketching.

ABSTRACT. The current study examined the effects of introductions on reading times and immediate and delayed recalls for brief scientific texts across two experiments (N = 219). The findings suggest that introductions improved participants’ memory for scientific explanations, both immediately and after a delay, without changes to reading times for scientific content. Both narrative and expository genres provided similar memorial benefits for scientific content, with narrative introductions also proving quite memorable for readers.

ABSTRACT. Answering questions from an available text and self-explaining target sentences while reading expository texts are well-documented learning activities to promote complex conceptual learning. This study compares the cognitive processes promoted by these two activities. We propose and test a moderated mediation model that includes two complementary pathways by which both activities have an impact on students’ deep learning. This model has theoretical and practical implications for conceptual learning.

ABSTRACT. Misconceptions interfere with learning and are difficult to correct. Two studies examined the interactive effects of constructed response prompt (self-explanation, think-aloud) and text type (refutational, non-refutational), and self-explanation training on students’ misconceptions regarding natural selection. In Study 1, students (N=240) were randomly assigned to prompt and text condition. In Study 2, students (N=153) were randomly assigned to prompt and training conditions. In both studies, vocabulary was the sole significant predictor of conceptions of natural selection.

ABSTRACT. We examined the relations of child characteristics (struggling reader status, language and cognitive skills), text features (narrative vs. expository genres), and nature of comprehension questions (literal vs. inferential comprehension questions) to discourse comprehension in oral language (listening comprehension). Data were from 523 English-speaking second graders. Genres and child characteristics explained performance variation in listening comprehension whereas nature of comprehension questions did not. Struggling reader status did not explain variation controlling for language and cognitive skills.

ABSTRACT. We examined how readers’ online affective and attentional experiences influenced comprehension after reading. Participants were periodically interrupted during reading to assess their affective valence (i.e., their feelings) and whether their minds had wandered away from the text. Results revealed that affective valence and mind-wandering influence levels of comprehension differently: wandering thoughts are overall negative for comprehension and positive valence negatively impacts shallow comprehension while increasing readers propensity to interpret emotion in a text.

ABSTRACT. Students may read and answer questions with the texts available (open-book) or unavailable (closed-book). No differences between both procedures for long-term retention are apparent, but the impact of question format has not been systematically investigated. We do it by recording online measures while reading and answering questions to provide evidences of processing. Answering open-ended questions produced more learning than answering multiple-choice questions in both procedures. Online data may shed light to explain this result.

ABSTRACT. This experiment employed a functional magnetic resonance imaging (fMRI) working memory task to examine how three sub-component processes: (1) Proactive Control, (2) Filtering, and (3) Disengagement contribute to a mechanistic explanation of the relation between working memory and reading skill. Results suggested that skilled readers deploy more prefrontal resources when cued proactively about task-relevant features than do less-skilled readers. In contrast, reading skill was not related to activation associated with attention-filtering or successful disengagement.

ABSTRACT. Readers’ understanding of a task guides processing decisions during reading, and higher task awareness should correlate with better task performance. Task awareness arises from a reader’s task model, but what supports task model construction? Strategies that support comprehension (paraphrasing, bridging and elaborative inferences) may influence performance indirectly by supporting task model construction. The goal of this study was to explore the hypothesis that task awareness partially mediates the relationship between comprehension strategies and literacy outcomes.

ABSTRACT. Today's expert readers do much more with texts than simply comprehending them: they synthesize multiple texts, critique authors' viewpoints, integrate multimodal documents. Setting off from a meta-analysis of contemporary accounts of such reading tasks, this paper proposes a new 'constructive reading' model to fill the gap between those tasks and the skills that make them possible: the construal of texts into abstract, idealized, purposeful renderings of their contents—as we do across the sciences.