The Recovery of Correlation Between Latent Abilities Using Compensatory and Noncompensatory Multidimensional IRT Models

This research examined correlation estimates between latent abilities when using the two-dimensional and three-dimensional compensatory and noncompensatory item response theory models. Simulation study results showed that the recovery of the latent correlation was best when the test contained 100% of simple structure items for all models and conditions. When a test measured weakly discriminated dimensions, it became harder to recover the latent correlation. Results also showed that increasing the sample size, test length, or using simpler models (i.e., two-parameter logistic rather than three-parameter logistic, compensatory rather than noncompensatory) could improve the recovery of latent correlation.     

Where do hypotheses come from?

Why are human inferences sometimes remarkably close to the Bayesian ideal and other times systematically biased? In particular, why do humans make near-rational inferences in some natural domains where the candidate hypotheses are explicitly available, whereas tasks in similar domains requiring the self-generation of hypotheses produce systematic deviations from rational inference. We propose that these deviations arise from algorithmic processes approximating Bayes’ rule. Specifically in our account, hypotheses are generated stochastically from a sampling process, such that the sampled hypotheses form a Monte Carlo approximation of the posterior. While this approximation will converge to the true posterior in the limit of infinite samples, we take a small number of samples as we expect that the number of samples humans take is limited. We show that this model recreates several well-documented experimental findings such as anchoring and adjustment, subadditivity, superadditivity, the crowd within as well as the self-generation effect, the weak evidence, and the dud alternative effects. We confirm the model’s prediction that superadditivity and subadditivity can be induced within the same paradigm by manipulating the unpacking and typicality of hypotheses. We also partially confirm our model’s prediction about the effect of time pressure and cognitive load on these effects.     

Testing Differential Item Functioning in Small Samples

Abstract

Differential item functioning (DIF) is a pernicious statistical issue that can mask true group differences on a target latent construct. A considerable amount of research has focused on evaluating methods for testing DIF, such as using likelihood ratio tests in item response theory (IRT). Most of this research has focused on the asymptotic properties of DIF testing, in part because many latent variable methods require large samples to obtain stable parameter estimates. Much less research has evaluated these methods in small sample sizes despite the fact that many social and behavioral scientists frequently encounter small samples in practice. In this article, we examine the extent to which model complexity—the number of model parameters estimated simultaneously—affects the recovery of DIF in small samples. We compare three models that vary in complexity: logistic regression with sum scores, the 1-parameter logistic IRT model, and the 2-parameter logistic IRT model. We expected that logistic regression with sum scores and the 1-parameter logistic IRT model would more accurately estimate DIF because these models yielded more stable estimates despite being misspecified. Indeed, a simulation study and empirical example of adolescent substance use show that, even when data are generated from / assumed to be a 2-parameter logistic IRT, using parsimonious models in small samples leads to more powerful tests of DIF while adequately controlling for Type I error. We also provide evidence for minimum sample sizes needed to detect DIF, and we evaluate whether applying corrections for multiple testing is advisable. Finally, we provide recommendations for applied researchers who conduct DIF analyses in small samples.     

Revisiting the past and back to the future: Horizons of cognition and emotion research

ABSTRACT

To commemorate that Cognition & Emotion was established three decades ago, we asked some distinguished scholars to reflect on past research on the interface of cognition and emotion and prospects for the future. The resulting papers form the Special Issue on Horizons in Cognition and Emotion Research. The contributions to Horizons cover both the field in general and a diversity of specific topics, including affective neuroscience, appraisal theory, automatic evaluation, embodied emotion, emotional disorders, emotion-linked attentional bias, emotion recognition, emotion regulation, lifespan development, motivation, and social emotions. We hope that Horizons will spark constructive debates, while offering guidance for the future growth and development of research on the interface between cognition and emotion. Finally, we provide an update on how Cognition & Emotion has fared over the past year, and announce some changes in editorial policies and the editorial board.     

An eye-tracking attention based model for abstractive text headline

Online network platforms provide great convenience for users to obtain information. However, it’s challenging to select the required information from enormous texts. Automatic text headline generation methods not only guide users to select the information they are interested in, but also solve the problem of information overload. Nevertheless, the existing works mainly utilize the grammar rules to obtain the key information of the source text, while ignoring the dwell time of user’s attention on different text contents. To address this issue, this paper proposes an abstractive text headline generation model based on the eye-tracking attention mechanism. Specifically, this model first relies on the eye-tracking data to establish the mapping relationship between text words and the words’ reading time. Then, an eye-tracking attention mechanism is constructed to judge the importance of different words. Finally, this attention mechanism is integrated into the encoder-decoder framework to generate a high-quality headline. Experimental results obtained from different datasets demonstrate that the headline generated by our model is more concise. Moreover, our proposed model outperforms significantly the classical headline generation models on ROUGE-1, ROUGE-2 and ROUGE-L.     

Opening the cuebox: The information children and young adults generate and rely on when making inferences from memory

We used a cue‐generation and a cue‐selection paradigm to investigate the cues children (9‐ to 12‐year‐olds) and young adults (17‐year‐olds) generate and select for a range of inferences from memory. We found that children generated more cues than young adults, who, when asked why they did not generate some particular cues, responded that they did not consider them relevant for the task at hand. On average, the cues generated by children were more perceptual but as informative as the cues generated by young adults. When asked to select the most informative of two cues, both children and young adults tended to choose a hidden (i.e., not perceptual) cue. Our results suggest a developmental change in the cuebox (i.e., the set of cues used to make inferences from memory): New cues are added to the cuebox as more cues are learned, and some old, perceptual cues, although informative, are replaced with hidden cues, which, by both children and young adults, are generally assumed to be more informative than perceptual cues.     

Lived religion and the religious field

ABSTRACT

In the study of lived religion, the focus on laypeople as religious agents can result in the simplistic juxtaposition of religion-as-practised by individuals and religion-as-prescribed by institutions. This perspective leads to analyses that over-emphasize agency and overlook the embeddedness of religious persons in intricate power relations that expand beyond the institution(s) closest to them. I propose that Pierre Bourdieu’s social theory, particularly as related to the religious field, offers tools for tackling this issue. While Bourdieu’s work has been criticized for relegating the laity to the status of passive consumers of religious goods, his theorizations can also be employed to produce nuanced micro-level accounts that prioritize laypeople’s practical knowledge of the field and the positions they take within it. Based on my case study of older Finnish women’s normative assessments related to religion, I demonstrate how scholars can investigate the role which their informants’ histories and investments within the religious field play in their religion-as-lived. The women in my study, lifelong members of Orthodox or Lutheran churches, defended their positions in the increasingly individualistic Finnish religious field through an emphasis on childhood socialization as the foundation of ‘proper’ religion.     

Using Linguistic Corpora as a Philosophical Tool

The central aims of this paper are to show how linguistic corpora have been used and can be used in philosophy and to argue that linguistic corpora and corpus analysis should be added to the philosopher’s toolkit of ways to address philosophical questions. A linguistic corpus is a curated collection of texts representing language use that can be queried to answer research questions. Among many other uses, linguistic corpora can help answer questions about the meaning of words and the structure of discourse. Through a discussion of examples, the paper shows that there are many philosophical questions that can be addressed by using a linguistic corpus. However, linguistic corpora need not (and often cannot) replace traditional philosophical methods. Lastly, it argues that the special properties of linguistic corpora, including their independence and the current ease and cheapness of access, make them an indispensable resource for philosophers.     

Effects of making errors in learning a foreign language

ABSTRACT

Kornell, Hays, and Bjork ([2009]. Unsuccessful retrieval attempts enhance subsequent learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35, 989–998) showed that incorrect guesses do not necessarily harm and might even improve the retention of information on a subsequent test. We sought to replicate the finding using educationally relevant stimuli. In two experiments, our participants either translated sentences in a foreign language receiving immediate feedback (errorful condition), or copied and studied the correct translation (errorless condition). After this training phase, a final test with the same sentences showed that translating sentences wrongly during training did not lower the accuracy of the errorful as compared to the errorless condition. Overall there was evidence that errorful training produced superior learning of the meaning and grammar of the foreign language sentences. The results support the idea that search processes activate a greater network of related knowledge in the errorful than in the errorless condition.     

Better together: a unified perspective on appraisal and emotion regulation

ABSTRACT

Advances in our understanding of appraisal processes and emotion regulation have been two of the most important contributions of research on cognition and emotion in recent decades. Interestingly, however, progress in these two areas has been less mutually informative than one might expect or desire. To help remedy this situation, we provide an integration of appraisal theory and the process model of emotion regulation by describing parallel, interacting and iterative systems for emotion generation and emotion regulation. Outputs of the emotion generation system are perceived by the emotion regulation system, and emotion regulation strategies then modulate emotion by intervening at specific stages of the emotion generation system, ultimately changing appraisal dimensions. We hope that our unified perspective will encourage and guide future research at the interface of cognition and emotion.