Behavior Research Methods最新文献

Mediation analysis, popular in many disciplines that rely on behavioral science data analysis techniques, is often conducted using ordinary least squares (OLS) regression analysis methods. Given that one of OLS regression's weaknesses is its susceptibility to estimation bias resulting from unaccounted-for random measurement error in variables on the right-hand sides of the equation, many published mediation analyses certainly contain some and perhaps substantial bias in the direct, indirect, and total effects. In this manuscript, we offer errors-in-variables (EIV) regression as an easy-to-use alternative when a researcher has reasonable estimates of the reliability of the variables in the analysis. In three real-data examples, we show that EIV regression-based mediation analysis produces estimates that are equivalent to those obtained using an alternative, more analytically complex approach that accounts for measurement error-single-indicator latent variable structural equation modeling-yet quite different from the results generated by standard OLS regression that ignores random measurement error. In a small-scale simulation, we also establish that EIV regression successfully recovers the parameters of a mediation model involving variables adulterated by random measurement error while OLS regression generates biased estimates. To facilitate the adoption of EIV regression, we describe an implementation in the PROCESS macro for SPSS, SAS, and R that we believe eliminates most any excuse one can conjure for not accounting for random measurement error when conducting a mediation analysis.

For decades, there has been a dearth of efficient tools for researchers to measure L2 Chinese proficiency. This study introduces ClozCHI, a cloze test developed to assess L2 Chinese proficiency across a range of levels from novice to advanced. Unlike existing Chinese cloze tests, ClozCHI comprises three passages with varying levels of difficulty. Its effectiveness was assessed with 225 L2 Chinese learners who participated in the Hanyu Shuiping Kaoshi (HSK) at Levels 3 to 6 and completed ClozCHI within 2 weeks before or after their HSK tests. Additionally, supplementary data were collected from 97 learners below HSK Level 3 without HSK testing. The psychometric analysis of the ClozCHI using both classical test theory (CTT) and item response theory (IRT) revealed that the test demonstrated appropriate difficulty, good discrimination, and high reliability from novice to advanced levels. ClozCHI scores showed strong correlations with HSK levels, demonstrating criterion-related validity. Confirmatory factor analysis (CFA) further supported its unidimensional structure. ClozCHI was more effective for assessing reading than listening or writing. These findings suggested that ClozCHI is a reliable and valid instrument for assessing L2 Chinese proficiency in research settings. ClozCHI is freely available for researchers from the Open Science Framework repository: https://osf.io/5kcrq/ .

Researchers in psychology traditionally use analysis of variance to examine differences between multiple groups or conditions. A less well-known, but valuable alternative is contrast analysis - a simple statistical method for testing directional, theoretically motivated hypotheses that are defined prior to data collection. In this article, we review the core concepts of contrast analysis for testing hypotheses in between-subjects and within-subjects designs. We also outline and demonstrate the largely unknown possibility of directly testing two competing contrasts against each other. In the tutorial part of the article, we show how such competing-contrast analyses can be conducted in the free, open-source software R using the package cofad. Because competing-contrast analysis is a straightforward, flexible, highly powered, and hypothesis-driven approach, it is a valuable tool to extend the understanding of cognitive and behavioral processes in psychological research.

Flanker tasks are among the most popular paradigms for studying conflict resolution, particularly the intuitive "arrows" flanker task. To date, there is no auditory equivalent of this task, potentially due to the risk of perceptual masking. Using the strategy of reducing masking through vocal differences between two competing talkers, the present study explores the validity of a speech-based flanker task with co-located talkers. One hundred and forty-two healthy young participants (aged 18-24) participated in three auditory flanker tasks using naturally-produced speech with increasing spectral separation. Stimuli were the command words 'left' and 'right,' as well as two neutral flankers designed to impose masking effects, one containing only acoustic-phonetic information, and another containing lexical information. On each trial, the target and flanker words were heard concurrently, with participants responding to the direction spoken by the target talker. Classic conflict effects were observed to the incongruent flanker, which could not be attributed to masking. Both masking and conflict effects were somewhat reduced by spectral separation. A final experiment with 62 healthy adults (aged 18-59) compared the test-retest reliability and distributional properties of this auditory directional command flanker task with the visual arrows flanker task. The auditory task demonstrated similar test-retest reliability and evolution of conflict effects over time to the visual task. This pattern of results suggests that the directional command flanker task is comparable to the arrow flanker task and can be utilized to quantify inhibitory control in the auditory modality.

Partially nested designs-where clustering occurs in some but not all study arms-are common in psychological and behavioral research. In these designs, clustering often arises in the treatment arm due to the treatment delivery, such as individuals clustered within treatment facilitators. This clustering structure raises substantive interest in exploring how treatment effects may vary across clusters. Yet, existing tools for estimating cluster-specific treatment effects in such designs remain limited, especially when clustering is treatment-induced, and the assignment of treatment arm individuals to clusters is nonrandomized. To address this gap and to make the estimation accessible, we introduce an R package, PND.heter.cluster, which implements methods described in Liu (2024) for estimating cluster-specific treatment effects in 2/1 partially nested designs with treatment-induced clustering. The package supports the use of machine learning methods to relax modeling assumptions and enhance estimation flexibility. This tutorial describes the main function of the package and illustrates its use through a step-by-step application example using data from a partially nested design evaluating a teacher professional development program.

Automated behavioral measurement using machine learning is gaining ground in psychological research. Automated approaches have the potential to reduce the labor and time associated with manual behavioral coding, and to enhance measurement objectivity; yet their application in young infants remains limited. We asked whether automated measurement can accurately identify newborn mouth opening-a facial gesture involved in infants' communication and expression-in videos of 29 newborns (age range, 9-29 days, 55.2% female, 58.6% White, 51.7% Hispanic/Latino) during neonatal imitation testing. We employed a three-dimensional cascade regression computer vision algorithm to automatically track and register newborn faces. The facial landmark coordinates of each frame were input into a support vector machine (SVM) classifier, trained to recognize the presence and absence of mouth openings at the frame level as identified by expert human coders. The SVM classifier was trained using leave-one-infant-out cross-validation (training: N = 22 newborns, 95 videos, 354,468 frames), and the best classifier showed an average validation accuracy of 75%. The final SVM classifier was tested on different newborns from the training set (testing: N = 7 newborns, 29 videos, 118,615 frames) and demonstrated 76% overall accuracy in identifying mouth opening. An intraclass correlation coefficient of .81 among the SVM classifier and human experts indicated that the SVM classifier was, on a practical level, reliable with experts in quantifying newborns' total rates of mouth opening across videos. Results highlight the potential of automated measurement approaches for objectively identifying the presence and absence of mouth opening in newborn infants.

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.