Showing papers on "Comprehension published in 2018"
TL;DR: A comprehensive tutorial review of the science of learning to read, spanning from children’s earliest alphabetic skills through to the fluent word recognition and skilled text comprehension characteristic of expert readers is presented.
Abstract: There is intense public interest in questions surrounding how children learn to read and how they can best be taught. Research in psychological science has provided answers to many of these questio...
447 citations
TL;DR: These findings demonstrate that, when successfully comprehending natural speech, the human brain responds to the contextual semantic content of each word in a relatively time-locked fashion.
286 citations
TL;DR: It was found that the shared variance between vocabulary, grammar, verbal working memory, and inference skills was a powerful longitudinal predictor of variations in both listening and reading comprehension.
Abstract: Listening comprehension and word decoding are the two major determinants of the development of reading comprehension. The relative importance of different language skills for the development of listening and reading comprehension remains unclear. In this 5-year longitudinal study, starting at age 7.5 years (n = 198), it was found that the shared variance between vocabulary, grammar, verbal working memory, and inference skills was a powerful longitudinal predictor of variations in both listening and reading comprehension. In line with the simple view of reading, listening comprehension, and word decoding, together with their interaction and curvilinear effects, explains almost all (96%) variation in early reading comprehension skills. Additionally, listening comprehension was a predictor of both the early and later growth of reading comprehension skills.
155 citations
01 Jun 2018
TL;DR: A methodology for reconstructing warrants systematically is developed and operationalized in a scalable crowdsourcing process, resulting in a freely licensed dataset with warrants for 2k authentic arguments from news comments.
Abstract: Reasoning is a crucial part of natural language argumentation. To comprehend an argument, one must analyze its warrant, which explains why its claim follows from its premises. As arguments are highly contextualized, warrants are usually presupposed and left implicit. Thus, the comprehension does not only require language understanding and logic skills, but also depends on common sense. In this paper we develop a methodology for reconstructing warrants systematically. We operationalize it in a scalable crowdsourcing process, resulting in a freely licensed dataset with warrants for 2k authentic arguments from news comments. On this basis, we present a new challenging task, the argument reasoning comprehension task. Given an argument with a claim and a premise, the goal is to choose the correct implicit warrant from two options. Both warrants are plausible and lexically close, but lead to contradicting claims. A solution to this task will define a substantial step towards automatic warrant reconstruction. However, experiments with several neural attention and language models reveal that current approaches do not suffice.
118 citations
TL;DR: The simple view of reading (SVR) as discussed by the authors proposes that performance in reading comprehension is the result of decoding and linguistic comprehension, and that each component is necessary but not sufficient for reading comprehension.
Abstract: The simple view of reading (SVR) proposes that performance in reading comprehension is the result of decoding and linguistic comprehension, and that each component is necessary but not sufficient f
109 citations
TL;DR: It is argued that adopting this framework means that neurobiological studies of language will be less focused on identifying correlations between brain activity patterns and mechanisms postulated by psycholinguistic theories, and increasingly more inclined towards integration of language with other cognitive systems.
100 citations
03 Oct 2018
TL;DR: The authors describes a general theory of comprehension, which is a comprehension model that allows for problem solving and planning when that becomes necessary to complement normal reading, but it leaves room for problem-solving and planning.
Abstract: This chapter describes a general theory of comprehension. The theory of text comprehension is a comprehension model in the sense, but it leaves room for problem solving and planning when that becomes necessary to complement normal reading. Theorists interested in text comprehension talk about the outcome of comprehension in terms of mental representations. The chapter focuses on the application of that theory to important research topics in the area of discourse comprehension: how words are identified in a discourse context, the representation of knowledge, the construction of macrostructures and situation models, and the role of inferences and working memory. It presents the open research questions, most pressingly about the formation of situation models and the modeling of macrostructures. The chapter shows how nontrivial comprehension assessment is. The levels of comprehension range from the superficial to the deep, from surface features to the textbase to the situation model.
90 citations
28 Aug 2018
TL;DR: This article investigated what makes questions easier across recent 12 MRC datasets with three question styles (answer extraction, description, and multiple choice) and employed simple heuristics to split each dataset into easy and hard subsets and examine the performance of two baseline models for each of the subsets.
Abstract: A challenge in creating a dataset for machine reading comprehension (MRC) is to collect questions that require a sophisticated understanding of language to answer beyond using superficial cues In this work, we investigate what makes questions easier across recent 12 MRC datasets with three question styles (answer extraction, description, and multiple choice) We propose to employ simple heuristics to split each dataset into easy and hard subsets and examine the performance of two baseline models for each of the subsets We then manually annotate questions sampled from each subset with both validity and requisite reasoning skills to investigate which skills explain the difference between easy and hard questions From this study, we observed that (i) the baseline performances for the hard subsets remarkably degrade compared to those of entire datasets, (ii) hard questions require knowledge inference and multiple-sentence reasoning in comparison with easy questions, and (iii) multiple-choice questions tend to require a broader range of reasoning skills than answer extraction and description questions These results suggest that one might overestimate recent advances in MRC
87 citations
TL;DR: It is convinced that such an AR MM system can be beneficially installed into anatomical education in order to prepare medical students more effectively for the clinical standards and for more interactive, student-centered learning.
Abstract: When preparing young medical students for clinical activity, it is indispensable to acquaint them with anatomical section images which enable them to use the clinical application of imaging methods. A new Augmented Reality Magic Mirror (AR MM) system, which provides the advantage of a novel, interactive learning tool in addition to a regular dissection course, was therefore tested and evaluated by 880 first-year medical students as part of the macroscopic anatomy course in 2015/16 at Ludwig-Maximilians-Universitat (LMU) in Munich. The system consists of an RGB-D sensor as a real-time tracking device, which enables the system to link a deposited section image to the projection of the user's body, as well as a large display mimicking a real-world physical mirror. Using gesture input, the users have the ability to interactively explore radiological images in different anatomical intersection planes. We designed a tutorial during which students worked with the system in groups of about 12 and evaluated the results. Subsequently, each participant was asked to assess the system's value by filling out a Likert-scale questionnaire. The respondents approved all statements which stressed the potential of the system to serve as an additional learning resource for anatomical education. In this case, emphasis was put on active learning, 3-dimensional understanding, and a better comprehension of the course of structures. We are convinced that such an AR MM system can be beneficially installed into anatomical education in order to prepare medical students more effectively for the clinical standards and for more interactive, student-centered learning.
84 citations
TL;DR: A large sample from the Human Connectome Project (HCP) dataset was analyzed and multivariate predictive models for these 2 skills using whole-brain gray matter volume features were built, showing that these models effectively captured individual differences in these 2 Skills and were able to significantly predict these components of reading comprehension for unseen individuals.
Abstract: Reading comprehension is a crucial reading skill for learning and putatively contains 2 key components: reading decoding and linguistic comprehension. Current understanding of the neural mechanism underlying these reading comprehension components is lacking, and whether and how neuroanatomical features can be used to predict these 2 skills remain largely unexplored. In the present study, we analyzed a large sample from the Human Connectome Project (HCP) dataset and successfully built multivariate predictive models for these 2 skills using whole-brain gray matter volume features. The results showed that these models effectively captured individual differences in these 2 skills and were able to significantly predict these components of reading comprehension for unseen individuals. The strict cross-validation using the HCP cohort and another independent cohort of children demonstrated the model generalizability. The identified gray matter regions contributing to the skill prediction consisted of a wide range of regions covering the putative reading, cerebellum, and subcortical systems. Interestingly, there were gender differences in the predictive models, with the female-specific model overestimating the males' abilities. Moreover, the identified contributing gray matter regions for the female-specific and male-specific models exhibited considerable differences, supporting a gender-dependent neuroanatomical substrate for reading comprehension.
73 citations
TL;DR: This paper conducted a year-long study of Quality Talk, a teacher-facilitated, small-group discussion approach designed to enhance students' ability to comprehend complex text, and found that students often struggle to understand complex text.
Abstract: Students often struggle to comprehend complex text. In response, we conducted an initial, year-long study of Quality Talk, a teacher-facilitated, small-group discussion approach designed to enhance...
TL;DR: In this article, the authors investigated the relationship between the level of text difficulty and reading fluency and reading comprehension in elementary students in the USA and found that increased text difficulty was negatively related to reading comprehension, although a few studies found no relationship.
Abstract: Prompted by the advent of new standards for increased text complexity in elementary classrooms in the USA, the current integrative review investigates the relationships between the level of text difficulty and elementary students’ reading fluency and reading comprehension. After application of content and methodological criteria, a total of 26 research studies were reviewed. Characteristics of the reviewed studies are reported including the different conceptualizations of text, reader, and task interactions. Regarding the relationships between text difficulty and reading fluency and comprehension, for students’ reading fluency, on average, increased text difficulty level was related to decreased reading fluency, with a small number of exceptions. For comprehension, on average, text difficulty level was negatively related to reading comprehension, although a few studies found no relationship. Text difficulty was widely conceptualized across studies and included characteristics particular to texts as well as relationships between readers and texts. Implications for theory, policy, curriculum, and instruction are discussed.
Posted Content•
TL;DR: This article propose a graph-based, language-guided attention mechanism for referring expression comprehension, which is composed of node attention component and edge attention component, and explicitly represents inter-object relationships, and properties with flexibility and power impossible with competing approaches.
Abstract: The task in referring expression comprehension is to localise the object instance in an image described by a referring expression phrased in natural language. As a language-to-vision matching task, the key to this problem is to learn a discriminative object feature that can adapt to the expression used. To avoid ambiguity, the expression normally tends to describe not only the properties of the referent itself, but also its relationships to its neighbourhood. To capture and exploit this important information we propose a graph-based, language-guided attention mechanism. Being composed of node attention component and edge attention component, the proposed graph attention mechanism explicitly represents inter-object relationships, and properties with a flexibility and power impossible with competing approaches. Furthermore, the proposed graph attention mechanism enables the comprehension decision to be visualisable and explainable. Experiments on three referring expression comprehension datasets show the advantage of the proposed approach.
01 Jan 2018
TL;DR: The present chapter aims to review the literature regarding the skills needed to master the affordances of advanced digital reading scenarios and suggests that children and adolescents possess some of these skills, but that students at all levels struggle in complex scenarios.
Abstract: The internet offers readers the unique opportunity to access rich information scenarios, but doing so requires the use of advanced digital reading skills. Examples of such scenarios are searching and acquiring information from multiple sources (e.g., hypertext, images, videos) and participating in the social exchange of information (e.g., web forums, social networks, commenting newspapers). In such scenarios, the reader has to cope with a) the constantly growing number of available information sources, b) the different formats in which digital information is presented, c) the varying quality of the information available. To deal with these affordances, individuals need to possess advanced reading skills that go beyond what is needed to understand a single text. Such skills include: a) search and navigation skills to select relevant web pages and hyperlinks and to avoid getting lost in hyperspace; b) integration of multiple pieces of information and multiple presentation formats (texts from different web pages, text and animations); and c) critical evaluation of information (e.g., assessing the trustworthiness of the information on a web page and evaluating the quality of a comment from a social network). Existing literature suggests that children and adolescents possess some of these skills, but that students at all levels struggle in complex scenarios. In the present chapter, we aim to review the literature regarding the skills needed to master the affordances of advanced digital reading scenarios.
TL;DR: This paper introduces a Span-Extraction dataset for Chinese machine reading comprehension to add language diversities in this area and hosted the Second Evaluation Workshop on Chinese Machine Reading Comprehension (CMRC 2018).
Abstract: Machine Reading Comprehension (MRC) has become enormously popular recently and has attracted a lot of attention. However, the existing reading comprehension datasets are mostly in English. In this paper, we introduce a Span-Extraction dataset for Chinese machine reading comprehension to add language diversities in this area. The dataset is composed by near 20,000 real questions annotated on Wikipedia paragraphs by human experts. We also annotated a challenge set which contains the questions that need comprehensive understanding and multi-sentence inference throughout the context. We present several baseline systems as well as anonymous submissions for demonstrating the difficulties in this dataset. With the release of the dataset, we hosted the Second Evaluation Workshop on Chinese Machine Reading Comprehension (CMRC 2018). We hope the release of the dataset could further accelerate the Chinese machine reading comprehension research. Resources are available: this https URL
TL;DR: The authors explored the role of vocabulary knowledge and morphological awareness in reading comprehension ability of Chinese as a heritage language (CHL) learners and found that vocabulary knowledge acquired in formal Chinese instruction contributed to morphology awareness and reading comprehension, to a greater extent than that gained through early exposure to spoken Chinese.
Abstract: This study explored the role of vocabulary knowledge and morphological awareness in reading comprehension ability of Chinese as a heritage language (CHL) learners. One hundred ninety five CHL students participated in this study and completed a series of measures including two sets of vocabulary knowledge (one consisting of items pertaining to early exposure to spoken Chinese and the other comprised of items selected from a pool of words in Chinese as a foreign language classrooms), morphological awareness (structural awareness and functional awareness), and reading comprehension ability (lexical inference and passage comprehension). Drawing upon structural equation modeling with a bootstrap estimation method, the study found that vocabulary knowledge and morphological awareness both contributed to reading comprehension among CHL learners. More critically, the results indicated that morphological awareness mediated the relation between vocabulary knowledge and reading comprehension in CHL learners. Furthermore, multiple regression analyses probed the relative contributions of vocabulary knowledge measures to morphological awareness and reading comprehension, and found that vocabulary knowledge acquired in formal Chinese instruction contributed to morphological awareness and reading comprehension, to a greater extent, than that gained through early exposure to spoken Chinese.
TL;DR: This article presented findings from an intervention across sixth and seventh grades to teach academic words to middle school students, investigating a progression of outcomes of the intervention across six and seven grades, and concluding that the intervention was effective.
Abstract: This article presents findings from an intervention across sixth and seventh grades to teach academic words to middle school students. The goals included investigating a progression of outcomes fro...
TL;DR: In this paper, the effects of linguistic (word decoding and comprehension-based print reading) and non-cognitive factors (reading frequency and self-efficacy) have extensively been studied for print reading; we know very little about their role in Internet reading, which is our focus in this study.
TL;DR: In this paper, the authors developed two psychometrically reliable tests, the S-CALS Instrument and the Spanish Academic Vocabulary (S-AVoc) Test, to explore the dimensionality of core academic-language proficiencies, as measured by these two tests; and to examine the contribution of the core academic language skills to reading comprehension.
Abstract: Although literacy achievement has improved in Chile, adolescents’ underperformance in reading comprehension is still a serious concern. In English, core academic-language skills (CALS) have been found to significantly predict reading comprehension, even controlling for academic vocabulary knowledge. CALS are high-utility language skills that support reading comprehension across school content areas. Guided by an operational definition of Spanish CALS (S-CALS), three goals drove this study: to develop two psychometrically reliable tests, the S-CALS Instrument and the Spanish Academic Vocabulary (S-AVoc) Test; to explore the dimensionality of core academic-language proficiencies, as measured by these two tests; and to examine the contribution of core academic-language proficiencies to reading comprehension. A cross-sectional sample of 810 Chilean students (grades 4–8) participated in four assessments that measured standardized reading comprehension, word-reading fluency, Spanish academic vocabulary, and S-CALS. Using classical test theory and item response theory analyses, results yielded robust reliability evidence for both instruments. Consistent with prior research, S-CALS and academic vocabulary scores displayed upward trends in higher grades yet considerable within-grade variability. Confirmatory factor analysis revealed that S-CALS and S-AVoc were best conceptualized as part of a higher order construct, the Spanish core academic-language and vocabulary skills (S-CALVS). The aggregated S-CALVS scores predicted reading comprehension, beyond the contribution of grade, school factors, and word-reading fluency. This study advances our scientific understanding of CALS as relevant for adolescent literacy beyond the English language. The high-utility school-relevant language and vocabulary skills offer promising tools to inform and evaluate innovative reading comprehension interventions for Spanish-speaking adolescents.
TL;DR: This study provides the first direct evidence that the availability of the speech production system is necessary for generating lexical prediction during sentence comprehension, and provides an explanation for the recruitment of language production during comprehension.
Abstract: Language comprehension often involves the generation of predictions. It has been hypothesized that such prediction-for-comprehension entails actual language production. Recent studies provided evidence that the production system is recruited during language comprehension, but the link between production and prediction during comprehension remains hypothetical. Here, we tested this hypothesis by comparing prediction during sentence comprehension (primary task) in participants having the production system either available or not (non-verbal versus verbal secondary task). In the primary task, sentences containing an expected or unexpected target noun-phrase were presented during electroencephalography recording. Prediction, measured as the magnitude of the N400 effect elicited by the article (expected versus unexpected), was hindered only when the production system was taxed during sentence context reading. The present study provides the first direct evidence that the availability of the speech production system is necessary for generating lexical prediction during sentence comprehension. Furthermore, these important results provide an explanation for the recruitment of language production during comprehension.
TL;DR: Results indicated that children with SCD had deficits in oral language, but these deficits were not as severe as their deficit in reading comprehension, which suggests that the oral language weaknesses of children withSCD represent a developmental delay rather than developmental deviance.
Abstract: The purpose of this meta-analysis was to examine the comprehension problems of children who have a specific reading comprehension deficit (SCD), which is characterized by poor reading comprehension despite adequate decoding. The meta-analysis included 86 studies of children with SCD who were assessed in reading comprehension and oral language (vocabulary, listening comprehension, storytelling ability, and semantic and syntactic knowledge). Results indicated that children with SCD had deficits in oral language (d = -0.78, 95% CI [-0.89, -0.68], but these deficits were not as severe as their deficit in reading comprehension (d = -2.78, 95% CI [-3.01, -2.54]). When compared to reading comprehension age-matched normal readers, the oral language skills of the two groups were comparable (d = 0.32, 95% CI [-0.49, 1.14]), which suggests that the oral language weaknesses of children with SCD represent a developmental delay rather than developmental deviance. Theoretical and practical implications of these findings are discussed.
Journal Article•
TL;DR: In this article, the authors review the literature on reading comprehension development and utilize cognitive frameworks to describe the knowledge, skills, and processes involved during reading for meaning, and highlight for educators how strategy instruction aligns directly with cognitive theories of reading comprehension.
Abstract: Increasingly, children enter the school system with a home language that differs from the language of the majority. Consequently, classrooms have students with diverse language backgrounds and teachers must develop reading comprehension instruction that meets the needs of all their students. To successfully plan instruction, it is critical for teachers to understand the strengths that second language learners (SLLs) bring to the classroom as well as the potential difficulties they face. Here we review the literature on reading comprehension development and utilize cognitive frameworks to describe the knowledge, skills, and processes involved during reading for meaning. We use these theories to explain why SLLs may have difficulty with reading comprehension and how we might leverage their strengths in domains such as executive control to support their reading comprehension development. We further highlight for educators how strategy instruction aligns directly with cognitive theories of reading comprehension. Ideally, such examples will enable educators to explicitly articulate for their students how effective strategies enable the development of comprehensive mental representations of the text, and ultimately enable good text comprehension.
17 Apr 2018
TL;DR: In this paper, the advantages and limitations of using thinking-out-loud (TOL) protocols for studying language comprehension are summarized in a convenient form and discussed in detail.
Abstract: The analysis of cognitive processes in real time is one of the most methodologically difficult tasks in all of psychology. This chapter summarizes in a convenient form the advantages and the limitations of using thinking-out-loud (TOL) protocols for studying language comprehension. There has been much controversy in the history of psychology about verbal reports as data, focusing primarily upon their oft-reported unreliability. In a detailed analysis of verbal reports as data, Ericsson and Simon clarified several points about their use that are important to keep in mind when thinking about TOL data. There are many kinds of TOL data that can be collected for studying text comprehension. Rumelhart and Graesser have both used a TOL task in which subjects answered specific questions after each sentence of a text. Rumelhart’s subjects answered five WH-questions (who, what, why, when, where) after each sentence. The texts began ambiguously, with each successive sentence providing additional constraints on what was going on.
TL;DR: It is concluded that children acquire word knowledge from incidental reading, that children with better comprehension skills are more efficient and competent learners, and that although varying the semantic diversity of the reading episodes did not improve learning per se in the laboratory manipulation of diversity, diversity does affect reading behavior in less direct ways.
TL;DR: Changes in neural activity in the alpha band are connected to episodes of mind‐wandering during listening comprehension, and in turn to decreased comprehension accuracy, demonstrating how alpha can be used to track attentional engagement during language comprehension and illustrating the dependence of successful language comprehension on attention.
TL;DR: This paper conducted a multi-site replication study with aspects of pre-registration in order to explore the feasibility of such an approach in second language (L2) research and found that no effects were detected in analyses conducted over all sites.
Abstract: We conducted a multi-site replication study with aspects of pre-registration in order to explore the feasibility of such an approach in second language (L2) research. To this end, we addressed open questions in a line of research that has examined whether having learners attend to form while reading or listening to a L2 passage interferes with comprehension (VanPatten, 1990; Leow, Hsieh, & Moreno, 2008). Our results are consistent with findings from the specific paradigm that we replicated (i.e., Leow et al., 2008) in that no effects were detected in analyses conducted over all sites. However, further investigation is warranted due to site-specific effects and methodological limitations. We found all aspects of the multi-site registered replication approach to be useful, although the registration component itself appears to be a more feasible and useful first step in order to increase the robustness and generalizability of findings in our field.
TL;DR: It is argued that media and reading trends in recent decades indicate broader social and cultural changes in which long-form deep reading traditionally associated with the printed book will be marginalised by prevailing media trends and the reading modes they inspire.
Abstract: This paper analyses major social shifts in reading by comparing publishing statistics with results of empirical research on reading. As media statistics suggest, the last five decades have seen two shifts: from textual to visual media, and with the advent of digital screens also from long-form to short-form texts. This was accompanied by new media-adequate reading modes: while long-form content invokes immersed and/or deep reading, we predominantly skim online social media. Empirical research on reading indicates that the reading substrate plays an important role in reading processes. For example, comprehension suffers when complex texts are read from screens. This paper argues that media and reading trends in recent decades indicate broader social and cultural changes in which long-form deep reading traditionally associated with the printed book will be marginalised by prevailing media trends and the reading modes they inspire. As these trends persist, it may be necessary to find new approaches to vocabulary and knowledge building.
TL;DR: Head-Toes-Knees-Shoulders performance predicted unique variance in children's Chinese reading comprehension concurrently beyond word reading and a set of linguistic comprehension skills, highlighting the important role of EF skills in beginning readers' reading comprehension.
Abstract: Background Word reading and linguistic comprehension skills are two crucial components in reading comprehension, according to the Simple View of Reading (SVR). Some researchers have posited that a third component should be involved in reading and understanding texts, namely executive function (EF) skills. Aim This study was novel in two ways. Not only did we tested EF skills as a predictor of reading comprehension in a non-alphabetic language (i.e., Chinese) to extend the theoretical model of SVR, we also examined reading comprehension further in kindergarten children (age 5) in Hong Kong, in the attempt to reveal possible early precursors of reading comprehension. Sample(s) A group of 170 K3 kindergarteners was recruited in Hong Kong. Methods Children's word reading was assessed. Their linguistic comprehension was assessed with phonological awareness, verbal short-term memory, and vocabulary knowledge. Using a structured observation task, Head-Toes-Knees-Shoulders (HTKS), we measured their composite scores for EF skills. Results Head-Toes-Knees-Shoulders performance predicted unique variance in children's Chinese reading comprehension concurrently beyond word reading and a set of linguistic comprehension skills. Conclusions The results highlight the important role of EF skills in beginning readers' reading comprehension.
TL;DR: The results of both studies highlight the advantages that eye-tracking provides when its use is accompanied by an appropriate bilingual sample, by stimuli that reflect actual bilingual language use, and by secondary tasks that do not invoke metalinguistic processes.
Abstract: Researchers who study code-switching using lab-based approaches face a series of methodological challenges; these include, but are not limited to, using adequate techniques and tasks that allow for processing that reflects real-language usage and selecting stimuli that reflect the participants’ code-switching community norms. We present two illustrative eye-tracking studies that consider these challenges. Study 1 tests whether experience with code-switching leads to differential processing of Spanish determiner-English noun code-switches (e.g., una cookie ‘a cookie’). Study 2 examines auxiliary-verb code-switches involving the progressive structure (e.g., estan cooking ‘are cooking’) and perfect structure (e.g., han cooked ‘have cooked’) while participants read either for comprehension or provide grammaticality judgments. The results of both studies highlight the advantages that eye-tracking provides when its use is accompanied by an appropriate bilingual sample, by stimuli that reflect actual bilingual language use, and by secondary tasks that do not invoke metalinguistic processes.