ABSTRACT. The purpose of this study is to synthesize existing studies on reading while listening to determine for whom and under what circumstances reading while listening is helpful for comprehension. A meta-analysis was conducted with 30 eligible studies (total N = 1945). A trivial benefit of reading while listening over reading only on comprehension was found g = .18, p = .01. However, this was limited to studies in which reading was paced by the experimenter.

ABSTRACT. This quantitative corpus study analyses graphic communication signals conveyed by the covers of contemporary romance novels in several sub-genres. It explores how such images encode information about physical attributes and attitudes desired by readers both for their corporeal selves and in their idealised mates, given divergent evolutionary incentives. By identifying recurring tropes, it seeks to engage with a broader discussion about cultural, societal, and evolutionary perspectives associated with body image and body size diversity.

ABSTRACT. In this experiment, we explored the effect of an emoji in an apology posted to a Microsoft Teams workplace chat on the perceived authenticity of the apology and subsequent likelihood of forgiveness for a workplace transgression. Results demonstrated that emotionality mediated the relationship between the emoji and authenticity. Message appropriateness had no effect on the relationship. Authenticity predicted likelihood of forgiveness.

ABSTRACT. This study examined the multimedia heuristic by comparing text vs. video, and the deception effect by comparing videos with and without decorative, irrelevant images, on metacomprehension judgments, in first-year college students, when tested in a remote, e-learning experimental scenario. Students judged their comprehension worse with video compared to text. Decorative images in video hindered performance but did not yield a deception effect. Participants’ judgments aligned better with their comprehension with text compared to videos.

ABSTRACT. This study examined the effect of reading instruction on second-language reading by measuring the recall rate of a narrative text with five situational dimensions. A total of 39 Japanese university students with varying English proficiencies participated in the within-subject design, where they read one text with and one without reading instruction. The results show that reading instructions improve recall rate, indicating that it can enhance second language readers’ comprehension and recall of the text.

ABSTRACT. In the current research, we investigated the breadth and depth of strategies that college-level second language readers employ when reading multiple academic texts in a digital environment and the extent to which these strategies and task-relevant navigation predict post-reading comprehension of the ideas, as measured by a multiple-choice test. The analyses indicated that both monitoring and evaluation strategies and task-relevant navigation were foundational for a better comprehension of academic ideas within and across multiple texts.

ABSTRACT. This study examined the effects of text centrality and relevance on EFL readers’ text memory. Participants read an expository text for general comprehension or to answer questions and then performed a written recall task. The results showed that EFL readers recalled text information in accordance with centrality. However, pre-reading questions induced readers to recall more peripheral information than central information. These results suggest that EFL readers’ text recall depends on both centrality and relevance.

ABSTRACT. We encounter narratives in text and visual media. Some research has argued that comprehension process including inference generation are important for media and operate similarly yet, few studies have explored this issue. The present study explored the extent to which there are advantages to generating bridging inferences in the context of narratives that are conveyed in either text or picture media. Implications for theories of comprehension will be discussed.

ABSTRACT. This study explored the impact of media on mental model construction for narratives. Participants thought aloud while reading text or picture versions of the same stories. Analyses of their constructed responses revealed differences regarding situation content that was mentioned in the response. Participants mentioned more mental state words when reading the text versions, whereas they mentioned more spatial place words in the picture versions. Implications for theories of comprehension will be discussed.

ABSTRACT. This paper uses syntactic indices to test variance in syntactic complexity and whether different profiles relate to ‘good’ and ‘bad’ writing. We conduct an examination on syntactic variance and its relationship to quality of writing. Findings show four syntactic profiles across good and bad scores. Qualitative characterization of the clusters shows variance in combinations of sentence complexity and word/phrase frequency across the different clusters. Overlap in syntactic profiles observed across scores.

ABSTRACT. The current project outlines pitfalls of bias that can affect the development and implementation of AWE algorithms: (1) the training datasets, (2) the practices of data aggregation and disaggregation, (3) analytical procedures involved in producing generalizations, (4) black box nature of AWE algorithms, (4) lack of AWE validation, and (6) lack of regulation. We propose preliminary solutions: (1) de-biasing algorithms, (2) creating responsive systems that allow for (3) iterative feedback, (4) transparency and (5) explainability.

ABSTRACT. Myside bias occurs when humans evaluate, generate or verify information for the benefit of their beliefs; these beliefs are emotionally charged value judgments. The present research (N = 390) tested whether processing-based or conviction-based measures reduce myside bias more effectively, both immediately after the measure and 14 days later. Although both measures had been supported by previous evidence, in the present experiment none of them was found to reduce myside bias at either time point.

ABSTRACT. This study explored the online cognitive processes and offline comprehension products readers generate across linguistic and non-linguistic visual narratives. Participants completed think-aloud and recall tasks with non-linguistic graphic and text versions of narratives. While reading, participants generated more backward-oriented inferences and emotion inferences for graphic narratives. For text, participants generated more forward-oriented inferences and goal statements. In recall responses post-reading, participants included more emotion inferences for graphic narratives but more accurate story information for text.

ABSTRACT. This study examined the role of the causal connectivity of the statements and the performance of elaboration question tasks in the comprehension of written or spoken materials about the prevention of gender-based violence. We asked a group of Ecuadorian college students to listen to or read the interview, perform an elaboration-question answering task, and to answer comprehension questions. Results indicated that establishing causal connections and elaborating on statements with high causal connectivity facilitate student learning.

ABSTRACT. This study was conducted to better understand the difficulties faced by struggling college readers, including those enrolled in developmental education programs (DE). Relations among proficiency in component reading skills, one’s propensity to engage reading strategies, and enrollment in developmental education courses were examined. Results indicated that vocabulary was a positive predictor of bridging and elaboration scores. DE enrollment was a negative predictor of elaboration scores, suggesting that DE readers were less likely to produce elaborations.

ABSTRACT. Social media has become a primary venue for individuals to consume and disseminate information through discourse. To examine the features of this discourse, we collected tweets across several contexts and calculated their linguistic features using Coh-Metrix. We then conducted a PCA to extract the components of Twitter discourse. Results illustrate that the tweets can be characterized in terms of the degree of persuasive language used, lexical sophistication and diversity, events and actions, and information density.

ABSTRACT. The ability to locate the place-on-the-page (POP) where information appeared in a text was first established in the 1970s. Since then, the widespread use of computers has led to changes in reading formats. This study extended prior work to demonstrate that topic knowledge improves not just comprehension of text, but also memory for where information appears. Further, while scrolling formats harmed comprehension, participants showed POP memory at above chance levels even in some digital contexts

ABSTRACT. We used a combination of text mining tools to analyze the content of an adult literacy-based intelligent tutoring system. We describe several word-level (frequency, morphology, orthographic characteristics, etc.), sentence-level (length, complexity, etc.), and passage-level characteristics (cohesion, readability, etc.) across lesson topics and difficulty branches. Our findings generally support the structure of the selected lessons and performance-based difficulty branches. Literacy learning engineers may find these and similar analyses useful in building and improving adaptive learning systems.

ABSTRACT. Many English speakers accept singular they if referent gender is irrelevant/unknown (e.g. a student forgot their backpack), but find they ungrammatical if gender is known (e.g. Lisa forgot their backpack), unlike speakers with innovative grammars who also accept this. We report a production study that tests what pronouns participants produce when writing about referents whose gender is known (e.g. Kate, Andy) vs. unknown but socially implied to be male (last-name-only style, e.g. Smith, Jones).

ABSTRACT. The clinical and research use of language sample analysis (LSA) for speakers of African American English (AAE) can be problematic in that the software programs used for linguistic analysis may potentially attribute features of the AAE dialect that are in contrast with MAE as “error codes” rather than “dialect features” or language differences. This investigation takes a step towards identifying typical patterns of conversational, narrative, and expository discourse among adolescents from primarily African American backgrounds.

ABSTRACT. The present article is an empirical study of a classroom training designed to improve vocational students’ sourcing skills and to examine the impact of this improvement on critical thinking, source memory, and reading comprehension. The training is an updated version of that of Braasch et al (2013). Data will be collected from a training group and compared to a control group in a pretest-posttest design.

ABSTRACT. We present detectors that automatically identify when students game the system, a maladaptive learning strategy where students attempt to succeed by exploiting properties of a learning environment. In contrast to previous detectors that detected this behavior within students’ interaction with learning activities, we detect within students’ text-based responses to open-ended questions. With 5-fold student-level cross-validation, the model reached an average AUC ROC of 0.815, demonstrating a reliable method for detecting gaming in open-ended questions.

ABSTRACT. The current study tested whether receiving a goal before study (through a test-expectancy manipulation) and being guided by adjunct questions (AQs) during study would assist readers in engaging in the appropriate cognitive processes necessary for the construction of a more complex situation model of the scientific phenomenon. The test-expectancy manipulation improved performance, but adding AQs did not. Several results suggest that AQs constrain the readers’ focus to the textbase.

ABSTRACT. While work on improving comprehension has primarily focused on open-ended generative activities, closed-ended practice tests using inference-type questions may also benefit understanding from text. Two studies were designed to test for differences between two different closed-ended practice test formats (multiple-choice and true-false). Multiple-choice practice tests consistently resulted in better comprehension outcomes compared to re-reading in the first study, and over true-false practice tests even when the text was available during practice testing in the second.

ABSTRACT. This study investigates the use of explicit instruction in both EFs and SRL by means of a question-processing (QAR) intervention, examining its impact on reading comprehension and on meta-strategic-awareness. The results indicate a significant improvement in reading comprehension and in meta-strategic awareness of the intervention group compared to the control group, pointing to the importance of engaging executive processes in reading comprehension interventions.

ABSTRACT. Grammatical aspect and temporal cues have been shown to impact discourse processing. In this preregistered study, we examined the impact of grammatical information (verb aspect and temporal cues) on mental model construction. Participants (n = 240) completed a sentence-completion task, wherein verb aspect (imperfective, perfective) and temporal cues (short vs long events, before vs after) were manipulated and the availability of target discourse concepts was assessed. Results are pending; however, predictions and hypotheses are discussed.

ABSTRACT. We will describe how a scenario-based assessment paradigm can be used to evaluate multiple document comprehension, including how it promotes multiple document processing, its instructional relevance, and how assessment forms can be used to track student growth in multiple document comprehension. As an example, we provide a comparison of single document comprehension assessment and multiple document comprehension assessment, both focusing on the same topic to show convergence and divergence of comprehension processes.

ABSTRACT. Prior beliefs influence comprehension and memory for multiple, conflicting texts. This study investigated the extent to which individual differences in reading comprehension skills attenuate the text-belief consistency effect. Undergraduates were administered a conceptual test and read four texts with conflicting information on homelessness. They were also assessed on their general prior knowledge and reading comprehension skill.

ABSTRACT. The Construction-Integration Model (Kintsch, 1988) establishes that prior knowledge predicts reading comprehension. Numerous empirical studies report the effect prior knowledge has on learning from text. However, the role that prior knowledge can have varies statistically. Prior knowledge is often analyzed as a control variable. However, prior knowledge can also be a moderator or predict in a mediation model. We conducted simulation studies with prior knowledge given its varied statistical effects. Methodological implications are considered.