ABSTRACT. Previous studies have shown the relations of working memory, perspective taking (i.e., theory of mind), and oral language to discourse comprehension. In the present study, we examined how these co-develop over time using longitudinal data from Kindergarten to Grade 4 (N=250). For many students, co-development was found whereas for students with low theory of mind in Kindergarten, growth was observed in working memory, vocabulary, and grammatical knowledge without commensurate growth in theory of mind.

ABSTRACT. Core Academic Language Skills (CALS) refer to the knowledge and deployment of high-utility language prevalent in school texts across content areas (e.g., complex syntax, logical connectives, anaphora). Operationalized through the CALS-Inventory, CALS contribute to reading comprehension (RC) for middle graders. In this study, we find 4th-graders’ CALS predicted unique variance in RC after accounting for register-general language skills (vocabulary, word recognition, and sentence comprehension), suggesting the utility of measuring both register-general and –specific discourse knowledge.

ABSTRACT. Inference making is fundamental to the construction of a mental model of a text, making valid and sensitive assessments of this skill essential for research on the development of reading comprehension. We present an oral assessment of inference making, suitable for pre-school to grade 3. The task has good construct validity (listening comprehension all grades; reading comprehension grades 1-3) and internal reliability (Cronbach’s alpha), and can be used successfully to predict longitudinal development of reading comprehension.

ABSTRACT. This paper reports results from a three-year measurement study designed to develop, refine and validate the Multiple-choice Online Causal Comprehension Assessment (MOCCA). Refinements made for the final version (e.g., revision processes), overall results (e.g., IRT), and how results inform directions for reading comprehension instruction will be discussed. Results suggest MOCCA will serve as a diagnostic measure of reading comprehension that identifies good/poor comprehenders, and types of poor comprehenders grounded in comprehension theory and basic research.

ABSTRACT. In this study we examined the efficacy of a fully-automated, interactive questioning intervention (TeLCI) designed to improve reading comprehension by training inference making in Grades 1 and 2 (n = 107). The effectiveness of questioning may differ depending on when questions are posed. An important consideration, and a question we will address in this presentation, is whether it is more beneficial to prompt the students during (i.e., online) or after (i.e., offline) the comprehension task.

ABSTRACT. The backfire effect in refutation texts describes the doubling down on misconceptions in the face of their refutation. We hypothesized that backfire effects would occur when no explanation of the connection was presented and would be less likely with explanations. We randomly assigned learners to read a refutation text with an explanation, no explanation, or filler information. Results indicated that learners decreased their misconceptions at post-test with the explanation condition statistically outperforming the filler text.

ABSTRACT. The current project examines people's appraisals of their ability to discount and ignore inaccuracies, and how this relates to their use of inaccurate text content. Previous work has indicated that individuals tend to overestimate what and how much they know. In line with those findings, people also overestimate their ability to discount inaccuracies. The degree to which people overestimate such discounting is predictive of the influence of inaccurate text content on related decisions and judgments.

ABSTRACT. We examined the extent to which readers' revised knowledge from reading a refutation text transfers to a subsequent transfer text that re-activates the misconception. Participants read refutation and transfer texts that addressed common misconceptions about vaccines. Reading time data show that under some circumstances, reading a refutation text prior to reading the transfer text reduces disruption readers experience while reading the transfer text, suggesting that transfer of the revised knowledge from refutation texts does occur.

ABSTRACT. We examined whether the use of personal anecdotes could promote knowledge revision. Participants read refutation-plus-explanation, refutation-only, and explanation-only texts that addressed misconceptions about vaccines and autism. Texts were either framed as personal anecdotes or traditional expository texts. The results showed that texts with anecdote features facilitated greater knowledge revision when readers were provided with an explanation in the refutation text. Additionally, participants experienced weaker knowledge revision when they read refutation-only texts, regardless of anecdotal features.

ABSTRACT. Research has consistently demonstrated that readers rely on inaccurate information from texts to complete post-reading tasks. We examined whether opportunities to independently fact-check information via on online search would reduce error rates on a general knowledge task following exposure to related inaccuracies. Overall, participants frequently took the opportunity to fact-check, utilizing online searches for more than a third of critical items. Moreover, fact-checking proved beneficial in reducing participants’ subsequent reproduction of inaccurate text content.

ABSTRACT. Text comprehension has many complexities that make it a challenge to study as well as to achieve. One way to simplify the complexity is to focus on the central role that a word has in reading: Its text-relevant meaning is the link between word identification and text comprehension. This perspective allows us to ask what happens while a word is being read and to seek evidence on how—if at all—the word is integrated into the reader’s understanding of the text. I review recent and ongoing ERP studies that provide such evidence, addressing the role of predictive vs. memory-based processes, the significance of sentence boundaries, and local vs global influences on meaning integration, and the influence of the reader’s task.  The results suggest a simplified account of the complex integrative and structure building processes that occur as a word is read.  

ABSTRACT. In the present study, we investigated the relations between the components of the simple view of reading (oral language and decoding), and two components of executive function (cognitive flexibility and working memory) in a sample of 272 native English-speaking 9- to 14.83-year-old children. Results of structural equation models indicated that both oral language and decoding fully mediated the relations between working memory and cognitive flexibility and reading comprehension. However, the degree to which each component of executive function influenced reading comprehension varied by oral language versus decoding, with cognitive flexibility contributing significantly more variance to oral language than decoding. These findings suggest that executive function likely influences reading comprehension through its relation with decoding and oral language, and provide additional support for the role of executive function in reading comprehension as a potentially crucial precursor to skilled reading.

ABSTRACT. To gain knowledge from texts, readers infer meaningful connections by activating and integrating incoming information. Focusing on elementary school children, we investigated the underlying processes and resulting knowledge representations. The results reveal individual and developmental differences, which shed light on factors that influence success and failure to learn from texts.

ABSTRACT. Skilled reading comprehension involves more than constructing a meaning representation from a single passage. Readers routinely need to search relevant texts, evaluate text information for accuracy and trustworthiness, and integrate potentially discrepant information across multiple texts. I will outline the RESOLV theory of purposeful reading (Britt, Rouet, & Durik, 2018), which aims to account for these processes. I will exemplify some of its mechanisms with data from developing and experienced readers.

ABSTRACT. Anne Britt to lead a discussion of the key themes and issues raised.

ABSTRACT. In the past decade we have seen a surge of interest in using neuroimaging to study text and discourse comprehension. In this symposium, we examine how new developments in neurocognitive research contribute to our understanding of reading comprehension, and to identify gaps that need to be investigated in the future. The four speakers/papers will each present findings based on recent work from the relevant labs that will address the pertinent research questions and challenges.

ABSTRACT. Behavioral studies have demonstrated differences between narrative and expository comprehension. However, few studies have examined the neural correlates of these different genres. In this study, children listened and read narrative and expository fMRI passages at four time points. Findings revealed that expository passages recruited the frontoparietal control network, a top-down, goal-oriented network, while narrative passages were more reliant on the default mode network, a network that is known to support social cognition.

ABSTRACT. Regional dialects represent the common linguistic code of many people, but there are rather few neuro¬linguistic studies investigating dialect comprehension on the text level. In an fMRI study we presented short fairy tales spoken in three German varieties to monolectal and bilectal speakers. The most prominent brain region differentially activated for speakers of the dialect was the left anterior temporal lobe (aTL), a region particularly important for language comprehension in context.

ABSTRACT. We combined eye-tracking and fMRI to investigate the interplay between visuo-orthographic and linguistic processes contributing to comprehension and oculomotor control. We investigated i.) the influence of visual multi-word information on oculomotor control and semantic integration, ii.) the influence of skipping as well as iii.) the effects of within-word refixations on haemodynamic responses and iv.) the hierarchical network dynamics underlying failed comprehension indexed by regression probabilities.

ABSTRACT. In this fMRI study, students read expository science texts for comprehension of basic STEM concepts while their BOLD responses and eye movements were recorded. High-skilled readers showed stronger connectivity among brain regions important for integrating semantic representation and building event representation. Graph-theoretical network measures of conceptual relations in the text also correlated with activity in inferior parietal and dorsolateral prefrontal regions, indicating text comprehension through semantic integration and working memory.

ABSTRACT. Forty-five participants read immigration-related arguments while their eye movements were recorded. Thinking aloud and evaluations of argument credibility, valence, and arousal were collected after each sentence. The preliminary analyses of the data indicate that the subjective ratings of argument credibility, valence and arousal correlate with eye movements as well as with the contents of the think-alouds. These findings imply that emotional reactions influence the processing strategies employed during argument reading.

ABSTRACT. Both study-related time pressure and hypertext reading have been found to independently increase cognitive load, resulting in weaker performance on tests of comprehension. Less is known about the effect of time pressure across different types of on-screen texts. The present study found that both time pressure and type of on-screen text significantly affected comprehension. Those learning under no time pressure, and from scrolling text, performed better than those in other text and time conditions.

ABSTRACT. The present study examined how the usefulness of source information to explain conflicting scientific claims affects readers’ attention to and use of this information. In an eye-tracking study, we presented 80 participants with two conflicting scientific claims with or without differences regarding the sources’ benevolence. Our results show that differences in benevolence between the sources (i.e., that source information can be used as conflict explanation) affect both readers’ attention to source information and conflict explanation.

ABSTRACT. Two studies tested whether the seductive effect of text comprehensibility that inclines low-knowledge readers to rely on their direct evaluation of scientific contents is prevented if readers can lean on evaluations of source credibility instead. The findings show that while source credibility affects content evaluation, it does not mitigate the influence of text comprehensibility. This indicates a lacking awareness of the relative importance of source evaluation in situations where prior topic knowledge is very low.

ABSTRACT. Readers’ evaluation (experiment 1, n=21) and memory (experiment 2, n=54) of sources as a function of their descriptions (physical aspect, knowledge) and the consistency of their claims (consistent, discrepant) was examined. Participants read 16 texts containing two embedded sources. In Experiment 1, discrepant claims and knowledge descriptions affected the evaluation of source credibility. In Experiment 2, memory of the link between the source and her/his knowledgeability was higher, but only when claims were discrepant.

ABSTRACT. Beliefs about the nature of knowledge (epistemic beliefs, EB) are supposed to be domain-specific and related to Multiple Document Comprehension (MDC). However, research so far has concentrated on the domain of science. In the present study, N=156 university students reported on their EB in either physics or history, and completed an MDC test. Latent class analyses revealed a class of more sophisticated EB in both domains, which was, however, not significantly related to MDC.

ABSTRACT. These two studies examined the relationship between students’ epistemologies and their ability to evaluate arguments. Study1 used domain-general and Study2 domain-specific epistemic questionnaires. Participants evaluated arguments and took an epistemological-belief survey before recalling arguments in a Surprise Recall task. Students’ general beliefs about the speed of knowledge-acquisition in Study1 and students’ science-specific beliefs about Justification through multiple-sources and through authority in Study 2 predicted how well they evaluated arguments. Implications of the findings are discussed.

ABSTRACT. We examine the association between epistemic beliefs (i.e., students' adoption of absolutist, multiplist, and evaluativist perspectives) and students' justifications for preferential text selection. Students justified their selection of the texts they preferred most according to content, evidence quality, author expertise, and author's provision of anecdotal evidence. Students selecting texts based on author's anecdotal evidence more strongly endorsed a multiplist perspective. Students justifying their text preferences based on evidence quality more strongly endorse an evaluativist perspective.

ABSTRACT. This study tested the onPAR model of academic literacy, which assumes that foundational literacy skills, metacognition, and motivation contribute to students’ performance on an academic literacy task. College students took a battery of general and situational assessments of foundational skills, metacognitive skills, and reading motivation. While only general literacy skills directly predicted academic literacy, the situational measures accounted for additional variance in the model. These results provide support for the onPAR model.  

ABSTRACT. Readers are generally unable to accurately monitor their comprehension when learning from text, meaning they are unable to distinguish what they have understood well from what they have not. This study investigated metacomprehension accuracy in Introductory Psychology. Accuracy at the start of the course was at zero. However, after students completed homework activities that prompted them to explain texts as they studied, and exposed them to example inference test questions, accuracy was improved.

ABSTRACT. We examined whether students were able to flexibly revise texts so they were appropriate for different audiences. Students (n=95) were instructed to revise multiple texts for two groups of readers: fourth-grade students or professors. Results indicated that students revised the texts appropriately for the audiences, but that the revisions interacted with their comprehension skills. Overall, this study suggests that reading skills can help students can engage in adaptive writing processes across multiple levels of text.

ABSTRACT. Dialogic discourse emphasizes making meaning through discussion, and stands in contrast to recitation, where the teacher initiates a question that is followed by a student response and is then evaluated by the teacher. Dialogic discussion is positively related to several academic outcomes, but its relationship to writing is less understood. Using a novel instrument that measures classroom talk, this paper investigates the relationship between dialogic discussion and writing in grades four through eight.

ABSTRACT. The effect of coherence marking on text comprehension was investigated by adding and removing causal, temporal, contrast and additive connectives in twenty authentic Dutch texts. 794 Dutch secondary school students each filled out four cloze tests (‘HyTeC-cloze’). Adding connectives only affected comprehension on a local level, not on a global level. The direction of the effect depended on the type of coherence relation. Contrast and causal connectives increased comprehension while additive connectives reduced comprehension.

ABSTRACT. We present the development of a standardized German writing test measuring fine-grained idea/plan-to-text composition skills (e.g. providing adequate information for the reader). We also present results on the psychometric properties of one scale (“adequate information content”). This scale forms one latent factor, has a high consistency, and achieves convergent and divergent validity. Furthermore, it is highly correlated with the quality of long student texts.

ABSTRACT. Although overhearers are at a comprehension disadvantage relative to addresses in well-understood conversations, how do they comprehend when interlocutors misunderstand? In this two phase study, pairs of strangers completed a difficult referential communication task and achieved varying levels of understanding. Recordings of these conversations were presented online to overhearers. Results demonstrated that overhearers understood references more accurately than addressees for poorly-understood dialogues, and that these overhearers' comprehension performance was predicted by their self-reported Perspective-Taking ability.

ABSTRACT. This study examines whether it is possible to promote source evaluation skills among primary school children. 85 children either received the self-developed intervention Clever Klicking or completed a control training. Results reveal that both groups increased their performance on tasks that required prompted or spontaneous evaluation of sources from written materials. However, greater learning gains were observed among students working with Clever Klicking, in particular with respect to spontaneous application of source evaluation skills.

ABSTRACT. A system of text simplification designed to address individual differences between poor readers (the TERENCE system) was tested with 194 7-11 year-old readers with different reading comprehension skills (good and average readers, poor readers and very poor readers). The simplification system offers to readers graded, cumulative, simplifications of the texts at 3 levels (global coherence, local coherence, linguistic simplification). The results show an interaction between levels of text simplification and readers’ reading comprehension skills.

ABSTRACT. Classroom learning activities often occurs through dialogue. We investigated classroom discussion patterns by considering group size in reading classes. We discovered there are much more speech utterance in small-group than in whole-class discussions, however, a large proportion of off-task utterance was observed during small-group sessions. Exchange of utterances tends to match with similar level of complexity with higher level utterances elicit more high level responses. Distinctive qualitative interaction patterns also observed between group sizes.

ABSTRACT. Increased focus on the language of schooling and its relationship to strong writing skills has led to calls for more instruction in academic language. This study of a new approach to teaching academic language deals with how students edit their writing to be more academic. The study finds that after a short intervention, students are able to identify elements of drafts and edit them to attempt to use more academic language.

ABSTRACT. We report that the hands posture (behind or in front oneself) during sentences learning determines memory performance for action language. Moreover, the EEG signatures of motor processes were modulated by action language but only in the front posture. During the recall, there was differential modulation of motor signatures for recalled and forgotten action sentences. We concluded that understanding manual-action sentences could involve simulations in the motor cortex, that impact memory performance.

ABSTRACT. This study examined the interactive effects of self-explanation training and constructed response prompt (self-explanation, think-aloud) on students’ misconceptions regarding natural selection. Undergraduate students (n=93) randomly assigned to training and prompt condition read a refutational text regarding natural selection and then completed the Conceptual Inventory of Natural Selection (CINS). A significant interaction between conditions revealed that benefits of self-explanation compared to think-aloud on CINS performance is only observed when students do not receive strategy training.

ABSTRACT. We used an adapted think-aloud protocol to assess to the content and temporal overlap in participants’ thoughts as they unfold in the absence of a deliberate task, i.e. when they are letting their minds wander. Results indicate that participant’s consecutive thoughts are related about half of the time (49.3%). An automated linguistic measure of cohesion across thoughts (constructed using auto-correlation) was also positively related to similarity in adjacent thoughts as well as future-orientated content.

ABSTRACT. In this study, we explore learners' discourse during online collaborative interactions by employing Group Communication Analysis, a methodology for quantifying and characterizing the discourse between learners in online interactions. GCA calculates six inter- and intra-personal sociocognitive measures, and from these identifies distinct interaction profiles through cluster analysis. We were able to diagnostically reveal six robust profiles of learners. This study presents a unique analysis of the sociocognitive processes that comprise the interaction between learners.

ABSTRACT. We investigated how the authority of a person asking students to read arguments affects their processing and recall for those arguments. It was found that the two sides of the argument (pro/con) were read in a less balanced way, and recall of claim was poorer when the requester was a non-authority vs an authority. These results suggest that reading behavior and memory are sensitive to aspects of the context in addition to the actual instructions.

ABSTRACT. We examined if individual differences in inference generation during reading in 10 year-old readers are reflected in their mental text representation. Children from three reading profiles based on inferences they generate; Literal- Paraphrasing-, and Elaborating Readers, read and recalled narrative texts. We examined whether central text units were recalled more often than peripheral, a centrality effect. All children showed a centrality effect, however, it was larger in Elaborating Readers who generate most inferences during reading.

ABSTRACT. Collaboration can be defined as two or more people working towards a common goal, in a contingent manner. Digital technology can militate against such collaboration. We compared pairs of adults using a software set-up that made their actions either contingent or non-contingent, to assess the effects on their languahe and behaviour. Pairs using the contingent set-up showed greater awareness of the other than those in the non-contingent version, more notably for behaviour than verbal interaction.

ABSTRACT. This study extends research on variation in sarcasm use by using Amazon Mechanical Turk to collect an age-diverse sample (ages 20-70) to address this gap in the literature. Across several measures of sarcasm use, the data support previous work demonstrating that males use, and report using, sarcasm more often than females. Additionally, these data show a negative relationship between sarcasm use and age, with older males being more sarcastic than older females.

ABSTRACT. We designed an experiment to test the role of elaborated feedback (EF) orienting learners toward the correct response (EFcorrect) or towards the mistake (EFmistake) when students have a second attempt to answer low-and high-level questions in learning science. Results show that students spend the same amount of time reading the feedback independently of their group, but EFcorrect group read more the messages after failure. Processing data are presented.

ABSTRACT. 30 adult pairs played a puzzle-solving computer game in a collaborative or non-collaborative version and then performed a communication task. Pairs playing the collaborative version collaborated more often, and all who collaborated showed an advantage in performance in the communication task, regardless of the version played. The results suggest that collaboration can be scaffolded by appropriate constraints in game design, and that such collaboration can have positive effects once players finish playing the game.

ABSTRACT. Previous studies have demonstrated that presenting scientific contents in a literary narrative format, compared to expository text, produces better learning. However, the precise cognitive processes that underlie this advantage have not been fully addressed. We will investigate how executive functions, theory of mind and other inference processes are involved. We wish to contribute to a better understanding of the cognitive processes involved in science learning across text formats and to inform learning practices.

ABSTRACT. In four experiments, influences of global-text structure (central theme) on word-to-text integration were examined. Results from ERP recordings in Exp1 and 2 indicate that local lexical-semantic binding processes dominate at sentence-initial words and were previously presented at ST&D. In Exp3 passages had a remote, instead of a local binding opportunity and effects of centrality appear. In Exp4, currently running, we stimulate global integration by asking participants to pick a title that best fits each passage.

ABSTRACT. The aim of the present experiment was to study how poor and average/good comprehenders (aged 9,5y.) resolve a pronominal reference when the resolution requires an inference about the agent of a verb and an elaboration of one of the two characters in the preceding sentence. We expected that poor comprehenders would have difficulties with the integration of several constraints and would be more prone to consider only one. This expectation was confirmed by the experiment.

ABSTRACT. In A Girl in the River: The Price of Forgiveness, the researchers have studied how language assigns different gender roles to men and women. The findings reveal that the documentary orbits around the theme of patriarchy and a sub-theme of women as the other or second sex. Furthermore, the documentary attempted to challenge the prevalent patriarchal structures in the society however, it paradoxically ends up reinforcing them by assigning more agency to men.

ABSTRACT. Our study experimentally examines the effects of a formula-scoring technique ("I don’t know"-answer option + penalties for incorrect answers) on the properties of multiple-choice text comprehension tests. 125 ninth and tenth grade students were randomly assigned to a formula-scoring or to an assessment-as-usual number-right condition. Students completed tasks on text comprehension, grammar, and orthography. Formula scoring seems to reduce the reliability of text-comprehension tests and to increase the recipients’ reliance on bottom-up-comprehension strategies.

ABSTRACT. Directives are utterances which function to elicit some kind of action or response on behalf of the listener. This study examines the use of directives by teachers for scaffolding the comprehension and communication of children with special education needs, using the 52,813 word SEN classrooms corpus and corpus linguistics-based query methods. Overall, results on the frequency of directives shed light upon the key functions of directives within scaffolding in SEN classrooms.

ABSTRACT. Validation of text information is an integral part of text comprehension. This self-paced reading experiment investigated combined effects of source credibility and plausibility of information during validation. Participants read short stories with a high-credible vs. low-credible person stating a true or false assertion. Reading times on target sentences were slower when a credible source stated a false assertion compared to a low credible source, implying that source information modulated the validation of implausible information.

ABSTRACT. We examined whether describing an event with imperfective or perfective verb aspect (Janet was taking/took a lemon to Carol) influenced 7-11 year olds’ judgement of the ongoingness of the event. Children listened to stories and answered questions to assess their perception of the event. They were more likely to state that imperfectively expressed events were ongoing than perfectively expressed events. Intervening text reduced the probability that imperfectively expressed events were considered ongoing for older children.

ABSTRACT. This study proposes a theory-centered perspective from which to examine the Theoretical Framework (TF) section of the research article. The proposed model – Creating a Theoretical Framework (CATF) – consists of three moves: Providing a theoretical background, Establishing a theoretical framework, and Sharpening the significance/focus of one’s study that uses the framework. This paper offers pedagogical suggestions regarding the teaching of the TF section.

ABSTRACT. In this study, a corpus of 120 essays was analyzed via Coh-Metrix to explore the differences between high- and low-knowledge learners’ essays concerning the essay type (i.e., argumentative vs. expository). The corpus was evaluated by an Automated Essay Scoring (AES) system. The results suggested that computational indices of cohesion (i.e., pronoun to noun phrase incidence score) can be reliable predictors of the evaluation of writing quality.

ABSTRACT. The purpose of the study was to evaluate the effectiveness of direct and discussion-based approaches to vocabulary instruction in sixth-grade science classes in an urban school. The vocabulary instruction took place in a multidisciplinary intervention, Dialogic Inquiry for Socioscientific and Conceptual Understanding in School Science (DISCUSS), designed to develop higher-order thinking skills and academic vocabulary in socio-scientific Issues-based curriculum. Students who engaged in direct vocabulary instruction and inquiry dialogue demonstrated a significant academic vocabulary gain.