ABSTRACT. We investigated comprehenders’ expectations about transitivity. Two transitivity-related properties were manipulated: aspect and clause type. In both experiments, participants read sentence fragments with optionally transitive verbs (“Bob was reading...”) and produced continuations. Exp.1 was conducted in English, Exp.2 in Brazilian Portuguese. Continuations were annotated for whether they contained an overt object. Main clauses yielded significantly more transitive continuations than embedded ones. Aspect also affected transitivity rate, but only when combined with a secondary cue.

ABSTRACT. We tested whether the connective because, shown to boost ratings of causal scientific explanations, also influences ratings of teleological explanations. Teleogical accounts do not illuminate causal relations between mechanisms and phenomena. Instead, they point to the purpose of phenomena. Since the quality of teleological explanations does not depend on a causal relation, it was predicted that the causal connective may not matter or even negatively influence ratings. The results are consistent with the second prediction.

ABSTRACT. Readers overlook some presupposed discrepancies (e.g., How many animals did MOSES take on the ark?). However, reading time inflation has recently indicated that readers detect both presupposed (given) and focused (new) text discrepancies. New experiments extended this finding to cleft construction target sentences (e.g., It was the boys that ate the ORANGES), which strongly distinguish between given and new sentence information. Reconciliation between satisfactory and deficient discourse validation is considered.

ABSTRACT. This study builds on the observation that while restrictive relative clauses (RCs) syntactically modify an NP, a coherence relation may be inferred between the RC and the matrix clause at the discourse level. By means of two continuation experiments, we demonstrate that both causal and concessive relations can hold between a restrictive RC and its matrix clause, and that these relations can influence expectations about the rest of the discourse, notably expectations about upcoming referents.

ABSTRACT. Examples and specifications occur frequently in text, but not much is known about how they function in discourse. We conducted a crowdsourced study to investigate how readers interpret these relations. The results show that they can indeed have two simultaneous functions: they can be used to illustrate/specify a situation and serve as an argument for a claim. We discuss the implications of these results and review the usability of crowdsourcing for answering such questions.

ABSTRACT. Although fantasy-text routinely violates general world knowledge, readers also hold fantasy information in memory. In four experiments, we explored the relation between fantasy-related contradictions and general world knowledge of fantasy. Across all experiments, the amount of fantasy context was manipulated. The overall finding was that as contextual support increased, the disruption in reading caused by fantasy-related contradictions decreased. Results are discussed within the context of the RI-Val model (O’Brien & Cook, 2016).

ABSTRACT. Participants read short narratives in which a character had a goal to achieve in a specific location. The character either made it to the location or not. Later in the narrative, after a variable amount of backgrounding after the location manipulation, a sentence was presented that was consistent or inconsistent with the earlier location. Reading times were longer on the inconsistent sentences. These results will be discussed with respect to memory-based and scenario-based text-processing theories.

ABSTRACT. This study examined the encoding and later recognition of characters making claims about an event (“sources”) as a function of the consistency or discrepancy of their statements. Other characters that made no claims were embedded in the stories. The source characters were better encoded and recognized in the discrepant than in the consistent condition. Eyetracking during the recognition task indicated that the source characters were more “tagged” together in participants’ memory in the discrepant condition.

ABSTRACT. Short narratives were constructed with an initial claim by a character followed later by a target sentence that was either consistent or inconsistent with that claim. When the narrator was neutral or credulous of the claim, inconsistent target sentences were read more slowly. However, when the narrator was skeptical of the claim, the consistency effect was eliminated. Our interpretation is that the resolution of the apparent inconsistency is mediated by the stance of the narrator.

ABSTRACT. We explored the effects of inconsistencies on subsequent comprehension. Participants read short narratives with familiar characters (e.g., Superman). An inconsistency (e.g., taking a cab to a robbery) reduced but did not fully eliminate the slow-down associated with a subsequent inconsistency. Additionally, an early inconsistency caused a slow-down for reading subsequent sentences containing information that was consistent with readers’ prior knowledge about the character (e.g., flying to a crime scene).

ABSTRACT. Participants read a series of social media statuses with embedded target statements containing accurate, neutral, or misleading general knowledge facts. Answers on a subsequent general knowledge quiz and sourcing questions were used to measure participants’ awareness and recall of target facts. Results replicate findings from narrative misinformation studies with more misinformed answers produced for misleading than neutral or accurate frames and more misinformed answers produced for less familiar compared to more familiar items.

ABSTRACT. To revise misconception about vaccines, three experimental text conditions were developed: refutation texts that identity, refute and explain misconceptions; refutation texts with emotional content (positive and negative); and non-refutation controls. Online (reading time) and offline (post-test) knowledge revision was measured. Refutation texts effectively revised vaccine misconceptions, and refutations with negative emotional content augmented this revision. Implications for persuasive science and health communication will be discussed.

ABSTRACT. People have been shown to use the false information they read even if they already possess accurate prior knowledge about those same ideas. We examined whether these effects differentially emerge based on political affiliation. We also assessed whether the opportunity to fact check information might attenuate any potential reliance on inaccuracies. Overall, individuals generally used inaccurate information, with fact checking emerging as broadly effective for reducing participants’ subsequent use of inaccurate information.

ABSTRACT. We explored whether inhibition was associated with the extent to which misconceptions are reactivation and disrupt comprehension. We found that inhibition may be necessary to reduce the interference of misconceptions during reading, but only when texts do not refute and explain the target misconceptions. When texts refute and explain the target misconceptions, the competing activation mechanism (proposed in KReC; Kendeou & O'Brien, 2014) may be sufficient to reduce the interference of misconceptions and facilitate revision.

ABSTRACT. Introductory psychology students received daily "Myth Busters" presentations that debunked misconceptions about psychology using the refutation method. For some myths, they took weekly quizzes to practice retrieving the correct information that refuted the misconception. The refutation method produced durable changes in students’ self-reported beliefs, but they often failed to retain the correct information. However, when students engaged in retrieval practice, they retained more correct information and were less likely to revert to producing the misconception.

ABSTRACT. This study investigated how the specificity of cues in task instructions affected processing of content related to a causal model afforded across texts. Adding a structural schema cue in task instructions increased memory and within-text integration during moment-to-moment processing more than a task providing only a semantic cue. However, the semantic cue alone was sufficient to increase integration in participants’ recall relative to a no-cue condition. Neither cue affected cross-text integration during moment-to-moment processing.

ABSTRACT. We examined whether proactive interference occurs during multiple text comprehension, and if readers can intentionally forget irrelevant information to reduce PI. Participants read 10 or 20 texts on a topic; some were instructed to intentionally forget prior-read texts between sets of 10 texts. PI and intentional forgetting conditions produced similarly high rates of memory intrusion errors from prior-read texts, far more than the no PI condition. We discuss implications of PI for multiple text comprehension.

ABSTRACT. Readers’ memory of sources as a function of their descriptions (appearance vs. knowledge), claim consistency (consistent vs. discrepant) and reading task (epistemic vs. perceptual evaluations) was examined. Fifty-eight undergraduates read 16 texts containing two embedded sources and performed a cued recall test. Discrepant claims and epistemic evaluations lead to better source recall. Descriptions of sources’ appearance lead to but better source recall in the epistemic task. Implications are discussed.

ABSTRACT. The purpose of this study was to examine the role played by metatextual knowledge on Internet Reading skills and how it develops on adolescence. 393 secondary students were tested on Metatextual Knowledge, Internet Reading skills and Internet Reading Engagement questionnaire, and a subsample was assessed on print reading skills. We found differences in metatextual knowledge among grades. Metatextual knowledge also predicted unique variance of Internet Reading scores beyond print reading skills.

ABSTRACT. Attention to and use of information about texts, such as author and publication venues, are referred to as sourcing and considered important skills in information societies. However, the term is not used consistently. To disclose differences, we sampled verbs from key-papers and identified four different dimensions of sourcing. A systematic review of 18 interventions revealed three clusters of studies relating to the dimensions. The results may be useful for embedding sourcing into study programs.  

ABSTRACT. Although a connection between engagement with stories and social cognitive outcomes has long been theorized, it is only recently that empirical investigations into this topic have begun to accumulate. Currently, there are a great number of studies exploring how stories and social cognition relate, across a wide range of research approaches. Here I propose a research framework that hopes to formalize how, when, and why engagement with stories might help to promote real-world social cognition. This framework, entitled Social Processes and Content Entrained by Narrative (SPaCEN), posits that stories might bolster social cognition either through (1) frequent engagement of social-cognitive processes, or (2) via the presentation of explicit content about social relations and the social world. Note that these two possibilities are not mutually exclusive and so both may occur. Based on this framework, I evaluate the extant evidence for the process and content accounts with the goal of directing future research toward clear gaps in the available evidence.

ABSTRACT. Over 80 studies have used relative metacomprehension accuracy measures to explore how readers monitor comprehension of text. The current study reports findings from a meta-analysis including: a postdiction superiority effect; higher accuracy for predictive judgments made on multiple texts rather than different sections of a single text , lower accuracy for single-item tests; and higher accuracy for memory-level rather than inference-level comprehension tests, suggesting that students default to memory-based cues when making judgments.

ABSTRACT. This study explored whether diagrams that typically illustrate key processes in biology textbooks might help or hinder comprehension monitoring. Using a standard paradigm, undergraduates read textbook excerpts on 6 topics either with or without diagrams; judged their understanding of each excerpt; and took comprehension tests. The diagrams were shown to improve comprehension. They also improved relative metacomprehension accuracy, and helped readers to use more valid cues as a basis for monitoring their own understanding.

ABSTRACT. The current study examined how cues to comprehension difficulty generated online during reading affected metacomprehension judgments and accuracy. Participants either read texts with or without inconsistencies. Results showed that participants based judgments on online cues when making judgments at the level of the overall text resulting in improved metacomprehension accuracy. Participants did not base judgments on online cues when making judgments at the level of a specific test items.

ABSTRACT. Two experiments examined the influence of information salience and psychological threat on metacognitive awareness of belief change. After reporting beliefs in a prescreening, subjects read about spanking effectiveness, reported beliefs, and recollected previous beliefs. Belief change predicted recollection accuracy. Threat was not elevated after reading a belief inconsistent text (Experiment 1), and value affirmation did not improve recollection (Experiment 2). Awareness of belief change appears to be poor and biased by salience of current beliefs.

ABSTRACT. We examined the relations between background knowledge, reading comprehension and students perceptions of their knowledge (metacognition). Students completed a background knowledge assessment that included an “I don’t know” option prior to reading and answering comprehension questions in three different purpose-driven multiple text assessments. The results indicated that reading comprehension and background knowledge were related, however, adding in the metacognitive judgment of knowledge increased the strength of the relationship with comprehension.

ABSTRACT. Two experiments were conducted to investigate how readers validate the plausibility of a narrative outcome based on the semantic features of an antecedent action and a character’s goal that elaborates the action. It was found that the ease to validate an outcome was facilitated by having 1) an explicitly stated goal that causally explains why the action would lead to the outcome; 2) a high degree of semantic overlap between the action and the outcome.

ABSTRACT. A reader of fantasy fiction can predictively infer outcomes that do not occur in the real world. I model an aspect of this process: the construction of a story-specific reference framework, which prescribes what can happen in that narrative universe. This structure is built from inferred information derived from GWK and the fictional context, which are integrated to alter generalized concepts beyond their original meanings. New processes are illustrated using techniques from ontological knowledge representation.

ABSTRACT. This study investigates the influence of L1 and L2 reading proficiency of EFL readers on English narrative text comprehension, focusing on temporal and spatial dimensions of the Event-Indexing model. Participants read stories for time and space in both L1 and L2 with time intervals and took L1 and L2 reading ability tests. Results indicated that only higher-level L2 readers tracked the temporal shift, yet they did not reach the level of situation model for space.

ABSTRACT. Several studies have shown that narratives can influence the self-concept. In the present study, our goal was to investigate whether stories portraying a strong protagonist can affect recipients’ own self-related control beliefs. Three experiments provide evidence for this assumption. The results suggest that this persuasive influence is mediated by transportation into the story, and that this effect, in turn, is mediated by the extent to which recipients experience emotions targeted by the narrative (event-congruent emotions).

ABSTRACT. Readers frequently encounter descriptions of characters’ voices in narratives, but do they encode and simulate this information? In two self-paced reading experiments, reading times were slower when the author of an email was described as a non-native, rather than a native, English speaker. This is consistent with the Auditory Perceptual Simulation (APS) account (Zhou & Christianson, 2016), which argues that readers mentally simulate characters’ voices, and activate auditory characteristics (e.g., speech rate) while reading.