Behavior Research Methods最新文献

Bayesian and structural-after-measurement (SAM) approaches have been developed, in part, to address limitations of conventional estimators in the context of structural equation models (SEMs) with latent interactions. Although both approaches have shown promise in a variety of contexts including small-sample studies, there is very little literature systematically comparing the relative benefits, limitations, and trade-offs among these approaches. In this study, we (a) compared the performance of estimators under each approach in multilevel SEMs with a within-, between-, or cross-level latent interaction and (b) demonstrated the flexibility of SAM approaches by extending and investigating them in partially nested SEMs with latent moderated mediation. The results suggest substantial differences between estimator performance as a function of the type of latent interaction. SAM approaches performed well with a variety of latent interactions in multilevel and partially nested SEMs, while Bayesian approaches, including those with informative priors, struggled with models that included a cross-level latent interaction and were not easily extended to partially nested SEMs. Overall, the results suggest that SAM approaches are a versatile and highly adaptable alternative or complement to conventional full-information estimators. To conclude, we outline estimator considerations based on the SEM type, latent interaction, and data structure.

To facilitate objective measures of proficiency for language users of diverse backgrounds, recent research in second language acquisition and multilingualism has developed short, yet reliable, tests of lexical knowledge in a wide range of languages. In this paper, we describe the development of LexKO, a brief lexically based test of Korean language proficiency, including its underlying logic, composition, intended use, and limitations. Three rounds of pilot and validation testing with first- and second-language Korean users resulted in a highly reliable Korean test comprising 60 items that can be completed in a few minutes. Freely available for other researchers to use, LexKO produces scores that correlate significantly with both first- and second-language Korean users' scores on a standardized proficiency test (an abridged version of the Test of Proficiency in Korean) and may thus be helpful in multi-part studies for obtaining a quick, valid measure of proficiency in Korean, one of the world's fastest-growing foreign languages.

In this work, I develop a mathematical formalization of the Replaced Elements Model (Wagner. Quarterly Journal of Experimental Psychology: Section B, 56(1), 7, 2003), within a general framework proposed by Ghirlanda (Journal of Mathematical Psychology, 64/65, 8-16, 2015, Journal of Mathematical Psychology, 85, 55-61, 2018), which provides a new way to apply and study the model. The result derived here has the novelty of explicitly stating how the model computes associative values without requiring either the application of complex algorithms or the use of special software. As a way of showing how to use this formalization, I apply it to the study of varied learning phenomena and several models, by either analytic means or simulations. In the process, I reproduce conclusions drawn previously for the Replaced Elements Model by other methods (Glautier. Behavior Research Methods, 39(4), 993-1000, 2007; Schultheis et al. Behavior Research Methods, 40, 435-441, 2008; Wagner. Experimental Psychology: Section B, 56(1), 7, 2003). As an interesting byproduct, I provide a general algorithm which may be applied to simulate the predictions of the replaced elements model, Rescorla-Wagner's model (Rescorla & Wagner. Classical conditioning, Current research and theory, 2, 64-69, 1972), and Pearce's configural model (Pearce. Psychological Review, 94(1), 61, 1994) among others. Concrete instances of the algorithm, coded in Python, are provided in the Appendix.

The recording and interpretation of lifting force data - such as load and grip forces - are central to studying weight perception. Typically, such data are collected using force transducer-embedded handles placed on top of objects. While effective, these handles may be impractical or undesirable for certain experimental paradigms. A potential alternative is a force transducer-embedded platform, but validation is needed to determine whether it captures force data with the same consistency and interpretability as the handle-based method, particularly given the potential for data loss around lift-off. In two experiments, we compared these methods by having participants lift light and heavy objects off a platform either directly or via handles to assess the convergent validity of experimental outcomes and the concurrent validity of the recorded data. Our findings indicate that the experimental outcomes and data from both methods were highly comparable, but only for the heavy objects. However, for the light object, platform-recorded force data showed lower agreement with handle-based measures, and several anticipated sensorimotor effects were not observed in the platform data. These discrepancies resulted in differences in experimental outcomes, particularly in the detection of switch effects, highlighting the platform's limitations for capturing lighter-weight interactions. Therefore, we suggest that while handles remain preferable for capturing rich force data, the platform method broadens methodological options and presents a viable and valid alternative.

Question-asking is a crucial aspect of human interaction. Questions fuel engagement, stimulate thought processes, foster learning, and facilitate information-seeking behavior. Yet, scarce empirical research exists on question-asking, or its relation to related cognitive capacities such as creativity, curiosity, and intelligence. The use of digital games as a research medium offers significant advantages for studying cognitive processes in natural settings. As such, this study empirically investigates how people ask questions in an online serious game. To do so, we developed an online game-Spot the Spy-where players are required to find a hidden spy amidst a crowded room, by asking a chatbot agent questions that guide them in their investigation. Our game thus offers to investigate question-asking in natural settings empirically, and optimal question-asking strategies, which we conducted in two studies. Study 1 (online, N = 103) focused on game development and exploratory validation, whereas Study 2 (in-lab, N = 100) focused on replication and extension. In both studies, participants completed a series of cognitive tasks assessing creativity, curiosity, and intelligence before playing the game. Our results highlight strategies related to optimal performance in the game, as well as how players' gameplay correlates with their cognitive abilities, especially with intelligence. Specifically, we found that higher intelligence scores were associated with more effective questioning strategies and better game performance. These insights highlight the potential of gamified environments to enhance our understanding of cognitive processes and advance the development of educational and training tools that foster strategic thinking and question-asking capacities.

Integrating artificial intelligence into psychological research represents a significant direction in contemporary psychology. Utilizing supervised and unsupervised machine learning techniques can further aid in understanding the nonlinear relationships of psychological concepts. In machine learning, variables, referred to as features, can encompass data from psychological scales, text, audio, and images. Current psychological research predominantly relies on frequentist approaches, where relationships between variables are typically based on regression, which often falls short in handling the nonlinear relationships of psychological characteristics. Therefore, we outline an innovative semi-automated workflow that empowers psychology researchers to leverage machine learning algorithms for intelligent model selection, facilitating the construction of more precise and insightful theoretical frameworks. This approach aims to achieve three primary research objectives: (1) automated hyperparameter tuning to attain optimal models; (2) identification of important features through interpretability techniques, facilitating feature selection based on calculated importance; (3) data-driven insights for theory building based on important features by integrating exploratory factor analysis with machine learning interpretability. In this paper, we provide an introduction to the basics of machine learning, describe the benefits of combining automated machine learning for researchers, and, using psychological resilience research as an example, offer a detailed annotated code workflow along with raw data. This low-code approach, designed with psychological research methodologies in mind, makes it highly accessible for psychological researchers.

We developed a new nostalgia induction using virtual reality. We compared this virtual reality task (VRT) with the established event reflection task (ERT) in terms of intensity of felt nostalgia and strength of psychological benefits produced by each induction method (Experiment 1) and the durability of these effects over time (Experiment 2). Offering initial validation for the VRT, Experiment 1 revealed that felt nostalgia and psychological benefits were higher in the nostalgia condition than in the control condition, irrespective of the induction method. In Experiment 2, we improved the VRT and measured felt nostalgia and psychological benefits at five time points, separated by 5-min intervals. The augmented VRT produced significantly stronger effects on felt nostalgia than did the ERT, and it retained this advantage over time. Compared to the ERT, the VRT did not produce significantly stronger effects on psychological benefits. Instead, psychological benefits were higher in the nostalgia than control condition at each time point except the final one, irrespective of induction method. Virtual reality environments can induce vivid, durable feelings of nostalgia and ensuing psychological benefits.

While numerous lexical databases provide rating norms for a wide range of words, resources for onomatopoeia remain scarce. Given the pivotal role of onomatopoeia in language development and its potential insights for the relationship between word phonology and word meaning, we introduce the Chinese Onomatopoeia Database (COD), comprising 97 one-character, 380 two-character, 91 three-character, and 183 four-character onomatopoeic words in Chinese (total N = 751). All words were rated by 311 native Chinese speakers for concreteness, imageability, context availability, age of acquisition (AoA), familiarity, semantic transparency, emotional valence, and emotional arousal. We demonstrated high reliability across these measures through Cronbach's alpha, split-half coefficients, and intra-class correlation coefficients (ICCs). Correlation analyses revealed significant associations among these lexical variables, including those between semantic and affective variables. Predictive validity of these variables was also examined using reaction times (RTs) and accuracy (ACC) obtained based on a lexical decision task, which showed that COD variables significantly predicted lexical decision RTs and ACC. Further analyses with two measures, Zipf and logCD, from the Chinese Children's Lexicon of Written Words (CCLOWW; Li et al., 2023) showed that these measures were significantly correlated with all COD variables. Even with the inclusion of CCLOWW-based Zipf or logCD measures in regression models, the COD-based variable still significantly predicted both RTs and ACC. The establishment of the COD not only fills a crucial gap in psycholinguistic resources but also provides a robust tool for future research into the cognitive and developmental underpinnings of language processing.

This work introduces TCBLex, a lexical database of Finnish literary works read by children between the ages of 7 and 15. We explain in detail the work done to build the corpus TCBLex is based on, including how books were sampled and collected, turned into text files, and finally processed. We also touch on legal considerations and how it is possible to build such a corpus in the EU. TCBLex contains over 11 million tokens that are annotated with parts-of-speech tags and lemmatized. We provide 14 different sub-lexicons in total, covering individual intended reading ages, age groups, as well as different genres. We also provide versions with additional morphological information, such as the cases and tenses of words. TCBLex provides various psycholinguistically interesting lexical statistics for both word types and lemmas, such as different frequency metrics, distributions, word lengths, numbers of syllables, morphological paradigm sizes, and for the first time in a Finnish lexicon, ages when words and lemmas are first encountered in books. TCBLex is freely available at https://doi.org/10.5281/zenodo.15655580 .

Correlation coefficients play a central role in scientific research to quantify the (linear) association between certain key variables of interest. Currently, hypothesis testing of correlation coefficients, such as whether a correlation equals zero or whether two correlations are equal, is mainly done using classical p values, despite their known limitations. An important cause of this problem is the limited availability of statistical software that supports alternative, Bayesian testing procedures. To address this shortcoming, the current tutorial paper showcases how to perform Bayesian hypothesis tests on correlation coefficients using the new BFpack module in the free and open-source software program JASP. The module supports Bayesian tests of various types of correlations such as product-moment correlations, polyserial correlations, or tetrachoric correlations, among others. Partial correlations can be tested by controlling for certain covariates. Moreover, both dependent and independent correlations can be tested to be zero or tested against each other. This tutorial aims to get researchers acquainted with this new flexible testing paradigm, which avoids the limitations of classical methods, and to make the methodology widely available to the research community.