Article
Social Sciences, Mathematical Methods
Matthew Naveiras, Sun-Joo Cho
Summary: In this study, empirical Bayes and hierarchical Bayes methods are proposed as alternatives to marginal maximum likelihood estimation (MMLE) for item parameter estimation in small sample sizes. Simulation results show that hierarchical Bayes methods can be acceptable alternatives to MMLE.
APPLIED PSYCHOLOGICAL MEASUREMENT
(2023)
Article
Psychology, Mathematical
John D. Patterson, Baptiste Barbot, James Lloyd-Cox, Roger E. Beaty
Summary: The visual modality is crucial for human creativity, but there is a lack of an effective automated solution for assessing visual creativity. This study introduces AuDrA, an automated drawing assessment platform, that can extract visual creativity scores from simple drawings and demonstrates its high correlation with human creativity ratings.
BEHAVIOR RESEARCH METHODS
(2023)
Article
Psychology, Mathematical
Kuan-Yu Jin, Thomas Eckes
Summary: Assessments using onscreen or internet-based technology for human ratings provide the benefit of automatically recording rating times. The hierarchical facets model for ratings and rating times (HFM-RT) is proposed to incorporate rating times as an additional data source to improve the quality of assessment outcomes. By analyzing simulated and real data, the HFM-RT successfully retrieved examinee and rater parameters, and demonstrated superior reliability indices in simulation. However, in the real-data analysis, the improvement in reliability was not significant due to the heterogeneity of examinees' writing proficiency.
BEHAVIOR RESEARCH METHODS
(2023)
Article
Social Sciences, Mathematical Methods
Jose H. Lozano, Javier Revuelta
Summary: The paper introduces a random weights linear logistic test model for measuring individual differences in operation-specific practice effects within a single test. The proposed model extends the linear logistic test model of learning and considers practice effects as random effects varying across examinees. Bayesian framework and simulation study were used to estimate and evaluate the model, which demonstrated good performance. An empirical study was conducted to illustrate the applicability of the model to real data, providing evidence of individual differences in operation-specific practice effects.
APPLIED PSYCHOLOGICAL MEASUREMENT
(2023)
Article
Psychology, Educational
Nataly Beribisky, Gregory R. Hancock
Summary: This study compared two fit indices, RMSEP which is a modified version of the adequacy index, and the RMSEP difference Delta RMSEP between nested models. The findings showed that RMSEP has increased sensitivity compared to Delta RMSEP as the number of indicator variables increases in the same model. The study also indicated that RMSEP has increased ability to detect noninvariance relative to Delta RMSEP in one-factor models.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2023)
Article
Mathematics, Interdisciplinary Applications
Lu Ou, Michael D. Hunter, Zhaohua Lu, Cynthia A. Stifter, Sy-Miin Chow
Summary: This study proposes an approach to fit mixed-effects continuous-time differential equation models, which are complex, possibly nonlinear and heterogeneous. The approach is evaluated using real data and Monte Carlo simulation, showing reasonable parameter estimates when identification constraints are met.
BRITISH JOURNAL OF MATHEMATICAL & STATISTICAL PSYCHOLOGY
(2023)
Article
Mathematics, Interdisciplinary Applications
Hung Tong, Cristina Tortora
Summary: Model-based clustering is a valuable approach to uncover heterogeneity in a data set. However, its use is limited when dealing with partially observed records. This article proposes a method that extends the mixture of multiple scaled contaminated normal distributions using the expectation-conditional maximization algorithm, allowing for robust parameter estimation and automatic outlier detection even with missing data.
JOURNAL OF CLASSIFICATION
(2023)
Article
Psychology, Educational
Christine E. Demars, Paulius Satkus
Summary: Marginal maximum likelihood is a popular estimation method in item response theory models, and Bayesian priors are often applied to likelihood in 3PL models when dealing with small sample sizes. Choosing appropriate priors for marginal maximum estimation has been overlooked. In this study, not using priors resulted in extreme and implausible parameter estimates for sample sizes of 1,000 or smaller. Applying priors to the c-parameters alleviated estimation problems for sample sizes of 500 or more, while both a-parameters and c-parameters needed priors for samples of 100. Bias was observed when the mode of the prior did not match the true parameter value, but the strength of the prior did not significantly affect the bias unless it was extremely informative. The root mean squared error (RMSE) of the a-parameters and b-parameters did not heavily depend on the mode or strength of the prior unless it was extremely informative. The RMSE of the c-parameters, similar to the bias, depended on the mode of the prior for c.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2023)
Review
Psychology, Mathematical
Emma M. Russell, Mitchell G. Longstaff, Heather Winskel
Summary: Researchers have been interested in the impact of individual differences on eyewitness memory, with the sex of the eyewitness being an important factor. There has been inconsistent findings regarding whether sex differences exist in eyewitness memory. However, the research suggests that neither males nor females have superior performance in the total amount of accurate information reported, but rather that females may have better memory for person-related details while males may perform better for details related to the surrounding environment. The own-gender bias was also consistently found.
PSYCHONOMIC BULLETIN & REVIEW
(2023)
Article
Psychology, Educational
Markus T. Jansen, Ralf Schulze
Summary: Thurstonian forced-choice modeling is a powerful tool for estimating item and person parameters and testing the model fit, aiming to reduce faking and other response tendencies in traditional self-report trait assessments. Recent developments have made it possible to calculate normative trait scores and enable comparisons between individuals using forced-choice assessment procedures. However, the use of item blocks in the multidimensional forced-choice format may yield biased results and should be approached with caution to reduce response tendencies.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2023)
Article
Psychology, Educational
Jianbin Fu, Xuan Tan, Patrick C. Kyllonen
Summary: This paper presents the application of the Rank-2PLM in forced-choice questionnaires and discusses the information functions for items and tests as well as the indicators related to trait score estimates.
JOURNAL OF EDUCATIONAL MEASUREMENT
(2023)
Article
Psychology, Mathematical
Hasibe Kahraman, Bianca de Wit, Elisabeth Beyersmann
Summary: This study explored cross-language morphological transfer mechanisms in highly proficient unbalanced Turkish-English bilinguals using a similar-script morphological translation priming paradigm. The results showed that the size of cross-language morphological transfer effects varied under different conditions, with stronger effects observed in bilinguals who acquired their second language at an earlier age. Additionally, the morphological priming effects were specifically influenced by the lexico-semantic relationship between the embedded word and its translation.
PSYCHONOMIC BULLETIN & REVIEW
(2023)
Article
Psychology, Mathematical
Felix Hao Wang, Meili Luo, Suiping Wang
Summary: This study examined the computational limit of word extraction and storage from continuous syllable sequences, finding that at least two occurrences of a word are required for successful segmentation, with four exposures generating more robust learning results.
PSYCHONOMIC BULLETIN & REVIEW
(2023)
Article
Psychology, Mathematical
Timothy K. Murphy, Nazbanou Nozari, Lori L. Holt
Summary: Communicating with a speaker with a different accent can affect one's own speech, and this effect is achieved through short-term statistical learning in passive listening and transfers to influence the listener's own speech production.
PSYCHONOMIC BULLETIN & REVIEW
(2023)
Editorial Material
Education & Educational Research
Steven Andrew Culpepper, Gongjun Xu
JOURNAL OF EDUCATIONAL AND BEHAVIORAL STATISTICS
(2023)
Article
Psychology, Mathematical
Hadeer Derawi, Casey L. Roark, Yafit Gabay
Summary: Speech communication relies on accurate perception and identification of speech sounds, which is achieved through categorization. Recent research suggests that learning new speech and non-speech categories depends on the procedural learning system. Individuals with Developmental Language Disorder (DLD) have difficulties in information-integration category learning but not in rule-based category learning.
PSYCHONOMIC BULLETIN & REVIEW
(2023)
Article
Psychology, Mathematical
Grant J. Taylor, Augustine T. Nguyen, Nathan J. Evans
Summary: This study validates the use of explicit double responding paradigms by assessing whether initial decisions of participants differ based on whether they were instructed that they could change their response. The results consistently show that allowing for changes of mind does not influence initial responses, with Bayesian analyses providing at least moderate evidence in favor of the null hypothesis.
PSYCHONOMIC BULLETIN & REVIEW
(2023)
Correction
Mathematics, Interdisciplinary Applications
C. W. Liu, R. P. Chalmers
BRITISH JOURNAL OF MATHEMATICAL & STATISTICAL PSYCHOLOGY
(2023)
Article
Psychology, Mathematical
Andrea Gregor de Varda, Marco Marelli, Simona Amenta
Summary: The study released a database of cloze probability values, predictability ratings, and computational estimates for a sample of English sentences. The analysis found that predictability ratings were the best predictors of EEG signals, self-paced reading times, and eye movement patterns when spillover effects were taken into account. The study also found that cloze probability estimates were the best predictors of early fixation patterns.
BEHAVIOR RESEARCH METHODS
(2023)
Article
Mathematics, Interdisciplinary Applications
Youmi Suk, Chan Park
Summary: This study proposes a framework for designing Optimal Treatment Regimes (OTRs) from multisite randomized trials in educational settings. The modifications to Q-learning and weighting methods improve their performance in handling hierarchical dependencies and nested structures. Simulation studies demonstrate the effectiveness of the proposed modifications in improving the OTRs' performance in multisite randomized trials.