Refine
Year of publication
Document Type
- Article (81)
- Postprint (16)
- Review (3)
- Monograph/Edited Volume (2)
- Part of Periodical (2)
- Other (1)
Keywords
- German (11)
- eye-tracking (9)
- interference (8)
- locality (8)
- Eye movements (7)
- sentence processing (7)
- Bayesian data analysis (6)
- Reading (6)
- individual differences (6)
- self-paced reading (6)
Institute
When researchers carry out a null hypothesis significance test, it is tempting to assume that a statistically significant result lowers Prob(H0), the probability of the null hypothesis being true. Technically, such a statement is meaningless for various reasons: e.g., the null hypothesis does not have a probability associated with it. However, it is possible to relax certain assumptions to compute the posterior probability Prob(H0) under repeated sampling. We show in a step-by-step guide that the intuitively appealing belief, that Prob(H0) is low when significant results have been obtained under repeated sampling, is in general incorrect and depends greatly on: (a) the prior probability of the null being true; (b) type-I error rate, (c) type-II error rate, and (d) replication of a result. Through step-by-step simulations using open-source code in the R System of Statistical Computing, we show that uncertainty about the null hypothesis being true often remains high despite a significant result. To help the reader develop intuitions about this common misconception, we provide a Shiny app (https://danielschad.shinyapps.io/probnull/). We expect that this tutorial will help researchers better understand and judge results from null hypothesis significance tests.
We present computational modeling results based on a self-paced reading study investigating number attraction effects in Eastern Armenian. We implement three novel computational models of agreement attraction in a Bayesian framework and compare their predictive fit to the data using k-fold cross-validation. We find that our data are better accounted for by an encoding-based model of agreement attraction, compared to a retrieval-based model. A novel methodological contribution of our study is the use of comprehension questions with open-ended responses, so that both misinterpretation of the number feature of the subject phrase and misassignment of the thematic subject role of the verb can be investigated at the same time. We find evidence for both types of misinterpretation in our study, sometimes in the same trial. However, the specific error patterns in our data are not fully consistent with any previously proposed model.
Previous studies have suggested that distinctive case marking on noun phrases reduces attraction effects in production, i.e., the tendency to produce a verb that agrees with a nonsubject noun. An important open question is whether attraction effects are modulated by case information in sentence comprehension. To address this question, we conducted three attraction experiments in Armenian, a language with a rich and productive case system. The experiments showed clear attraction effects, and they also revealed an overall role of case marking such that participants showed faster response and reading times when the nouns in the sentence had different case. However, we found little indication that distinctive case marking modulated attraction effects. We present a theoretical proposal of how case and number information may be used differentially during agreement licensing in comprehension. More generally, this work sheds light on the nature of the retrieval cues deployed when completing morphosyntactic dependencies.
Experiments in research on memory, language, and in other areas of cognitive science are increasingly being analyzed using Bayesian methods. This has been facilitated by the development of probabilistic programming languages such as Stan, and easily accessible front-end packages such as brms. The utility of Bayesian methods, however, ultimately depends on the relevance of the Bayesian model, in particular whether or not it accurately captures the structure of the data and the data analyst's domain expertise. Even with powerful software, the analyst is responsible for verifying the utility of their model. To demonstrate this point, we introduce a principled Bayesian workflow (Betancourt, 2018) to cognitive science. Using a concrete working example, we describe basic questions one should ask about the model: prior predictive checks, computational faithfulness, model sensitivity, and posterior predictive checks. The running example for demonstrating the workflow is data on reading times with a linguistic manipulation of object versus subject relative clause sentences. This principled Bayesian workflow also demonstrates how to use domain knowledge to inform prior distributions. It provides guidelines and checks for valid data analysis, avoiding overfitting complex models to noise, and capturing relevant data structure in a probabilistic model. Given the increasing use of Bayesian methods, we aim to discuss how these methods can be properly employed to obtain robust answers to scientific questions.
Previous research has found that comprehenders sometimes predict information that is grammatically unlicensed by sentence constraints. An open question is why such grammatically unlicensed predictions occur. We examined the possibility that unlicensed predictions arise in situations of information conflict, for instance when comprehenders try to predict upcoming words while simultaneously building dependencies with previously encountered elements in memory.
German possessive pronouns are a good testing ground for this hypothesis because they encode two grammatically distinct agreement dependencies: a retrospective one between the possessive and its previously mentioned referent, and a prospective one between the possessive and its following nominal head. In two visual world eye-tracking experiments, we estimated the onset of predictive effects in participants' fixations.
The results showed that the retrospective dependency affected resolution of the prospective dependency by shifting the onset of predictive effects.
We attribute this effect to an interaction between predictive and memory retrieval processes.
In the picture-word interference paradigm, participants name pictures while ignoring a written or spoken distractor word. Naming times to the pictures are slowed down by the presence of the distractor word. The present study investigates in detail the impact of distractor and target word properties on picture naming times, building on the seminal study by Miozzo and Caramazza. We report the results of several Bayesian meta-analyses based on 26 datasets. These analyses provide estimates of effect sizes and their precision for several variables and their interactions. They show the reliability of the distractor frequency effect on picture naming latencies (latencies decrease as the frequency of the distractor increases) and demonstrate for the first time the impact of distractor length, with longer naming latencies for trials with longer distractors. Moreover, distractor frequency interacts with target word frequency to predict picture naming latencies. The methodological and theoretical implications of these findings are discussed.
Intuitively, strongly constraining contexts should lead to stronger probabilistic representations of sentences in memory. Encountering unexpected words could therefore be expected to trigger costlier shifts in these representations than expected words. However, psycholinguistic measures commonly used to study probabilistic processing, such as the N400 event-related potential (ERP) component, are sensitive to word predictability but not to contextual constraint. Some research suggests that constraint-related processing cost may be measurable via an ERP positivity following the N400, known as the anterior post-N400 positivity (PNP). The PNP is argued to reflect update of a sentence representation and to be distinct from the posterior P600, which reflects conflict detection and reanalysis. However, constraint-related PNP findings are inconsistent. We sought to conceptually replicate Federmeier et al. (2007) and Kuperberg et al. (2020), who observed that the PNP, but not the N400 or the P600, was affected by constraint at unexpected but plausible words. Using a pre-registered design and statistical approach maximising power, we demonstrated a dissociated effect of predictability and constraint: strong evidence for predictability but not constraint in the N400 window, and strong evidence for constraint but not predictability in the later window. However, the constraint effect was consistent with a P600 and not a PNP, suggesting increased conflict between a strong representation and unexpected input rather than greater update of the representation. We conclude that either a simple strong/weak constraint design is not always sufficient to elicit the PNP, or that previous PNP constraint findings could be an artifact of smaller sample size.
Inferences about hypotheses are ubiquitous in the cognitive sciences. Bayes factors provide one general way to compare different hypotheses by their compatibility with the observed data. Those quantifications can then also be used to choose between hypotheses. While Bayes factors provide an immediate approach to hypothesis testing, they are highly sensitive to details of the data/model assumptions and it's unclear whether the details of the computational implementation (such as bridge sampling) are unbiased for complex analyses. Hem, we study how Bayes factors misbehave under different conditions. This includes a study of errors in the estimation of Bayes factors; the first-ever use of simulation-based calibration to test the accuracy and bias of Bayes factor estimates using bridge sampling; a study of the stability of Bayes factors against different MCMC draws and sampling variation in the data; and a look at the variability of decisions based on Bayes factors using a utility function. We outline a Bayes factor workflow that researchers can use to study whether Bayes factors are robust for their individual analysis. Reproducible code is available from haps://osf.io/y354c/. <br /> Translational Abstract <br /> In psychology and related areas, scientific hypotheses are commonly tested by asking questions like "is [some] effect present or absent." Such hypothesis testing is most often carried out using frequentist null hypothesis significance testing (NIIST). The NHST procedure is very simple: It usually returns a p-value, which is then used to make binary decisions like "the effect is present/abscnt." For example, it is common to see studies in the media that draw simplistic conclusions like "coffee causes cancer," or "coffee reduces the chances of geuing cancer." However, a powerful and more nuanced alternative approach exists: Bayes factors. Bayes factors have many advantages over NHST. However, for the complex statistical models that arc commonly used for data analysis today, computing Bayes factors is not at all a simple matter. In this article, we discuss the main complexities associated with computing Bayes factors. This is the first article to provide a detailed workflow for understanding and computing Bayes factors in complex statistical models. The article provides a statistically more nuanced way to think about hypothesis testing than the overly simplistic tendency to declare effects as being "present" or "absent".
Factorial experiments in research on memory, language, and in other areas are often analyzed using analysis of variance (ANOVA). However, for effects with more than one numerator degrees of freedom, e.g., for experimental factors with more than two levels, the ANOVA omnibus F-test is not informative about the source of a main effect or interaction. Because researchers typically have specific hypotheses about which condition means differ from each other, a priori contrasts (i.e., comparisons planned before the sample means are known) between specific conditions or combinations of conditions are the appropriate way to represent such hypotheses in the statistical model. Many researchers have pointed out that contrasts should be "tested instead of, rather than as a supplement to, the ordinary 'omnibus' F test" (Hays, 1973, p. 601). In this tutorial, we explain the mathematics underlying different kinds of contrasts (i.e., treatment, sum, repeated, polynomial, custom, nested, interaction contrasts), discuss their properties, and demonstrate how they are applied in the R System for Statistical Computing (R Core Team, 2018). In this context, we explain the generalized inverse which is needed to compute the coefficients for contrasts that test hypotheses that are not covered by the default set of contrasts. A detailed understanding of contrast coding is crucial for successful and correct specification in linear models (including linear mixed models). Contrasts defined a priori yield far more useful confirmatory tests of experimental hypotheses than standard omnibus F-tests. Reproducible code is available from https://osf.io/7ukf6/.
Parsing costs as predictors of reading difficulty : an evaluation using the Potsdam Sentence Corpus
(2008)
There is a wealth of evidence showing that increasing the distance between an argument and its head leads to more processing effort, namely, locality effects: these are usually associated with constraints in working memory (DLT: Gibson, 2000: activation-based model: Lewis and Vasishth, 2005). In SOV languages, however, the opposite effect has been found: antilocality (see discussion in Levy et al., 2013). Antilocality effects can be explained by the expectation based approach as proposed by Levy (2008) or by the activation-based model of sentence processing as proposed by Lewis and Vasishth (2005). We report an eye-tracking and a self-paced reading study with sentences in Spanish together with measures of individual differences to examine the distinction between expectation- and memory based accounts, and within memory-based accounts the further distinction between DLT and the activation-based model. The experiments show that (i) antilocality effects as predicted by the expectation account appear only for high-capacity readers; (ii) increasing dependency length by interposing material that modifies the head of the dependency (the verb) produces stronger facilitation than increasing dependency length with material that does not modify the head; this is in agreement with the activation-based model but not with the expectation account; and (iii) a possible outcome of memory load on low-capacity readers is the increase in regressive saccades (locality effects as predicted by memory-based accounts) or, surprisingly, a speedup in the self-paced reading task; the latter consistent with good-enough parsing (Ferreira et al., 2002). In sum, the study suggests that individual differences in working memory capacity play a role in dependency resolution, and that some of the aspects of dependency resolution can be best explained with the activation-based model together with a prediction component.
Scanpaths have played an important role in classic research on reading behavior. Nevertheless, they have largely been neglected in later research perhaps due to a lack of suitable analytical tools. Recently, von der Malsburg and Vasishth (2011) proposed a new measure for quantifying differences between scanpaths and demonstrated that this measure can recover effects that were missed with the traditional eyetracking measures. However, the sentences used in that study were difficult to process and scanpath effects accordingly strong. The purpose of the present study was to test the validity, sensitivity, and scope of applicability of the scanpath measure, using simple sentences that are typically read from left to right. We derived predictions for the regularity of scanpaths from the literature on oculomotor control, sentence processing, and cognitive aging and tested these predictions using the scanpath measure and a large database of eye movements. All predictions were confirmed: Sentences with short words and syntactically more difficult sentences elicited more irregular scanpaths. Also, older readers produced more irregular scanpaths than younger readers. In addition, we found an effect that was not reported earlier: Syntax had a smaller influence on the eye movements of older readers than on those of young readers. We discuss this interaction of syntactic parsing cost with age in terms of shifts in processing strategies and a decline of executive control as readers age. Overall, our results demonstrate the validity and sensitivity of the scanpath measure and thus establish it as a productive and versatile tool for reading research.
Linear mixed-effects models have increasingly replaced mixed-model analyses of variance for statistical inference in factorial psycholinguistic experiments. Although LMMs have many advantages over ANOVA, like ANOVAs, setting them up for data analysis also requires some care. One simple option, when numerically possible, is to fit the full variance covariance structure of random effects (the maximal model; Barr, Levy, Scheepers & Tily, 2013), presumably to keep Type I error down to the nominal a in the presence of random effects. Although it is true that fitting a model with only random intercepts may lead to higher Type I error, fitting a maximal model also has a cost: it can lead to a significant loss of power. We demonstrate this with simulations and suggest that for typical psychological and psycholinguistic data, higher power is achieved without inflating Type I error rate if a model selection criterion is used to select a random effect structure that is supported by the data. (C) 2017 The Authors. Published by Elsevier Inc.
There is a wealth of evidence showing that increasing the distance between an argument and its head leads to more processing effort, namely, locality effects; these are usually associated with constraints in working memory (DLT: Gibson, 2000; activation-based model: Lewis and Vasishth, 2005). In SOV languages, however, the opposite effect has been found: antilocality (see discussion in Levy et al., 2013). Antilocality effects can be explained by the expectation-based approach as proposed by Levy (2008) or by the activation-based model of sentence processing as proposed by Lewis and Vasishth (2005). We report an eye-tracking and a self-paced reading study with sentences in Spanish together with measures of individual differences to examine the distinction between expectation- and memory-based accounts, and within memory-based accounts the further distinction between DLT and the activation-based model. The experiments show that (i) antilocality effects as predicted by the expectation account appear only for high-capacity readers; (ii) increasing dependency length by interposing material that modifies the head of the dependency (the verb) produces stronger facilitation than increasing dependency length with material that does not modify the head; this is in agreement with the activation-based model but not with the expectation account; and (iii) a possible outcome of memory load on low-capacity readers is the increase in regressive saccades (locality effects as predicted by memory-based accounts) or, surprisingly, a speedup in the self-paced reading task; the latter consistent with good-enough parsing (Ferreira et al., 2002). In sum, the study suggests that individual differences in working memory capacity play a role in dependency resolution, and that some of the aspects of dependency resolution can be best explained with the activation-based model together with a prediction component.
Parsing costs as predictors of reading difficulty: An evaluation using the Potsdam Sentence Corpus
(2008)
The surprisal of a word on a probabilistic grammar constitutes a promising complexity metric for human sentence comprehension difficulty. Using two different grammar types, surprisal is shown to have an effect on fixation durations and regression probabilities in a sample of German readers’ eye movements, the Potsdam Sentence Corpus. A linear mixed-effects model was used to quantify the effect of surprisal while taking into account unigram and bigram frequency, word length, and empirically-derived word predictability; the so-called “early” and “late” measures of processing difficulty both showed an effect of surprisal. Surprisal is also shown to have a small but statistically non-significant effect on empirically-derived predictability itself. This work thus demonstrates the importance of including parsing costs as a predictor of comprehension difficulty in models of reading, and suggests that a simple identification of syntactic parsing costs with early measures and late measures with durations of post-syntactic events may be difficult to uphold.
Eye fixation durations during normal reading correlate with processing difficulty but the specific cognitive mechanisms reflected in these measures are not well understood. This study finds support in German readers’ eyefixations for two distinct difficulty metrics: surprisal, which reflects the change in probabilities across syntactic analyses as new words are integrated, and retrieval, which quantifies comprehension difficulty in terms of working memory constraints. We examine the predictions of both metrics using a family of dependency parsers indexed by an upper limit on the number of candidate syntactic analyses they retain at successive words. Surprisal models all fixation measures and regression probability. By contrast, retrieval does not model any measure in serial processing. As more candidate analyses are considered in parallel at each word, retrieval can account for the same measures as surprisal. This pattern suggests an important role for ranked parallelism in theories of sentence comprehension.
There is a wealth of evidence showing that increasing the distance between an argument and its head leads to more processing effort, namely, locality effects; these are usually associated with constraints in working memory (DLT: Gibson, 2000; activation-based model: Lewis and Vasishth, 2005). In SOV languages, however, the opposite effect has been found: antilocality (see discussion in Levy et al., 2013). Antilocality effects can be explained by the expectation-based approach as proposed by Levy (2008) or by the activation-based model of sentence processing as proposed by Lewis and Vasishth (2005). We report an eye-tracking and a self-paced reading study with sentences in Spanish together with measures of individual differences to examine the distinction between expectation- and memory-based accounts, and within memory-based accounts the further distinction between DLT and the activation-based model. The experiments show that (i) antilocality effects as predicted by the expectation account appear only for high-capacity readers; (ii) increasing dependency length by interposing material that modifies the head of the dependency (the verb) produces stronger facilitation than increasing dependency length with material that does not modify the head; this is in agreement with the activation-based model but not with the expectation account; and (iii) a possible outcome of memory load on low-capacity readers is the increase in regressive saccades (locality effects as predicted by memory-based accounts) or, surprisingly, a speedup in the self-paced reading task; the latter consistent with good-enough parsing (Ferreira et al., 2002). In sum, the study suggests that individual differences in working memory capacity play a role in dependency resolution, and that some of the aspects of dependency resolution can be best explained with the activation-based model together with a prediction component.