Article
Psychology, Educational
Goran Pavlov, Alberto Maydeu-Olivares, Dexin Shi
Summary: This study examines the accuracy of p-values using MV correction to the distribution of SRMR in SEM models. It found that the MV-corrected SRMR statistic provides accurate Type I errors under normality, but the accuracy decreases in models with excess kurtosis. In non-normal data, the MV-corrected LR test seems to outperform the MV-corrected SRMR.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2021)
Article
Mathematics, Interdisciplinary Applications
Dexin Shi, Christine DiStefano, Alberto Maydeu-Olivares, Taehun Lee
Summary: Research has shown that the performance of root mean square error of approximation (RMSEA) is suboptimal for assessing structural equation models with small degrees of freedom (df). This study compares the performance of standardized root mean square residual (SRMR) and comparative fit index (CFI) in small df models with different factor loadings, sample sizes, and model misspecifications. The results indicate that in comparison to RMSEA, SRMR and CFI are less influenced by df and can provide more useful information in differentiating models with varying degrees of misfit. Researchers are advised to exercise caution when interpreting RMSEA for models with small df and rely more on SRMR and CFI.
MULTIVARIATE BEHAVIORAL RESEARCH
(2022)
Article
Psychology, Multidisciplinary
Victoria Savalei, Jordan C. Brace, Rachel T. Fouladi
Summary: Comparison of nested models in structural equation modeling (SEM) often involves the use of fit indices and chi-square difference tests. This article argues that comparing RMSEA values by simply taking their difference is problematic and suggests the use of RMSEA(D) associated with the chi-square difference test instead. The authors highlight the need to modify current practices, especially in the assessment of measurement invariance.
PSYCHOLOGICAL METHODS
(2023)
Article
Psychology, Mathematical
Suzanne Jak, Terrence D. Jorgensen, Mathilde G. E. Verdam, Frans J. Oort, Louise Elffers
Summary: This tutorial explains how to conduct power calculations for SEM using power4SEM without using Monte Carlo methods. By focusing on model fit instead of statistical significance, power4SEM provides two computationally efficient methods for power calculations of SEM.
BEHAVIOR RESEARCH METHODS
(2021)
Article
Materials Science, Multidisciplinary
Showkat Ahmad Lone, Tabassum Naz Sindhu, Anum Shafiq, Fahd Jarad
Summary: This study proposes a new statistical model using a new alpha power transformation and the Gumbel Type-II distribution, and demonstrates its effectiveness and better fit compared to other models through simulation and real-life data analysis.
RESULTS IN PHYSICS
(2022)
Article
Mathematics, Interdisciplinary Applications
Nataly Beribisky, Robert A. Cribbie
Summary: It is suggested to use equivalence testing for evaluating model fit in SEM. This study proposes novel variations of equivalence tests based on popular fit indices. Monte Carlo simulations show that equivalence tests in SEM have good error control and power for detecting well-fitting models. They also provide unique interpretational advantages compared to other fit evaluation methods.
BRITISH JOURNAL OF MATHEMATICAL & STATISTICAL PSYCHOLOGY
(2023)
Article
Engineering, Multidisciplinary
Hedong Li, Yaozhi Luo, Demi Ai
Summary: In this study, a compressive sampling matching pursuit (CoSaMP) algorithm was developed for restoring electromechanical admittance (EMA) signatures-based damage identification. Validation was conducted through tests on artificial crack identification for a standardized concrete cube, and the proposed approach showed promising availability for damage identification using EMA technique encountered with lossy data.
Article
Biochemistry & Molecular Biology
Satyamvada Maurya, Amita Jain, Md Tabish Rehman, Ali Hakamy, Farkad Bantun, Mohamed F. AlAjmi, Vineeta Singh, Aafreen Zehra, Feroz Khan, Shafiul Haque, Bhartendu Nath Mishra
Summary: This study aimed to identify potential new drug targets/candidates for multi-drug-resistant Mycobacterium tuberculosis (Mtb) through computational techniques. The protein structure of ribonuclease VapC2 of Mtb H37Rv was analyzed, and docking studies were performed with 22 anti-TB drugs. Rifampicin showed the best binding energy and was selected as the parent molecule for designing derivatives. One of the derivatives, RIF-155841, showed activity in inhibiting VapC2. Molecular dynamics simulations confirmed the stability of the VapC2-RIF155841 complex, and a chemical synthesis scheme was designed for future validation.
Article
Mathematics, Applied
Pham Huu Anh Ngoc
Summary: This paper introduces explicit criteria for the contraction of stochastic differential equations, extending the well-known criterion for contraction of solutions of deterministic differential equations. An illustrative example is provided in the paper.
COMMUNICATIONS IN NONLINEAR SCIENCE AND NUMERICAL SIMULATION
(2021)
Article
Chemistry, Multidisciplinary
Jing Chen, Xiaoquan Lu
Summary: A maximum information extraction method based on Wavelet Transform (WT) is proposed in this paper for resolving the complex NIR signal. The results demonstrate that the method serves as an effective tool for obtaining the maximum valuable information in NIR study.
FRONTIERS IN CHEMISTRY
(2022)
Article
Virology
Nasrullah Khan, Asma Arshad, Muhammad Azam, Ali Hussein Al-marshadi, Muhammad Aslam
Summary: The COVID-19 pandemic has had a significant impact globally, leading researchers and governments to seek ways to control the spread of the virus and accelerate the treatment process. To effectively predict the development trend of the pandemic and reduce losses, they have developed a forecasting model to help researchers, governments, and others devise response strategies.
JOURNAL OF MEDICAL VIROLOGY
(2022)
Article
Mathematics
Showkat Ahmad Lone, Tabassum Naz Sindhu, Marwa K. H. Hassan, Tahani A. A. Abushal, Sadia Anwar, Anum Shafiq
Summary: Statistical models are crucial for data analysis, and researchers are always looking for potential or the latest statistical models to fit different data sets. In this investigation, a T-X transformation and the Gumbel Type-II model were used to create an improved statistical model. Simulation evaluation was conducted to assess the efficacy of the parameters. Two actual data sets were utilized to demonstrate the application of the T-X approach in producing the new and improved Gumbel Type-II (NIGT-II) distribution, and the results suggest that the NIGT-II distribution outperforms the Gumbel Type-II distribution.
Article
Mathematics, Applied
Anmin Mao, Shuyao Lu
Summary: In this paper, we investigate a Schrodinger equation with combined Hartree type and square-root nonlinearities. We prove the existence of the normalized ground state under certain assumptions, extending and complementing previous research.
MEDITERRANEAN JOURNAL OF MATHEMATICS
(2023)
Article
Quantum Science & Technology
Giorgio Celebre
Summary: The concept of Quantum Noise is fascinating and has attracted increasing interest from both fundamental and applied perspectives. The development of advanced technologies for practical quantum experiments and potential applications in quantum cryptography has further fueled this interest. This paper presents a heuristic approach, as an alternative to existing methods, to evaluate the Root Mean Square Quantum Noise, which quantifies the impact of measuring apparatuses on noise in quantum measurements.
QUANTUM INFORMATION PROCESSING
(2022)
Article
Computer Science, Interdisciplinary Applications
Pinaki Ranjan Mohanty
Summary: This paper studies a new sixth-order compact discretization method for mildly nonlinear differential equations. The three-step AGE iteration method is also discussed as a powerful numerical computation device. Error analysis and performance comparison are conducted.
ENGINEERING WITH COMPUTERS
(2022)
Article
Mathematics, Interdisciplinary Applications
Dexin Shi, Christine DiStefano, Alberto Maydeu-Olivares, Taehun Lee
Summary: Research has shown that the performance of root mean square error of approximation (RMSEA) is suboptimal for assessing structural equation models with small degrees of freedom (df). This study compares the performance of standardized root mean square residual (SRMR) and comparative fit index (CFI) in small df models with different factor loadings, sample sizes, and model misspecifications. The results indicate that in comparison to RMSEA, SRMR and CFI are less influenced by df and can provide more useful information in differentiating models with varying degrees of misfit. Researchers are advised to exercise caution when interpreting RMSEA for models with small df and rely more on SRMR and CFI.
MULTIVARIATE BEHAVIORAL RESEARCH
(2022)
Article
Psychology, Educational
Zhehan Jiang, Mark Raymond, Christine DiStefano, Dexin Shi, Ren Liu, Junhua Sun
Summary: Computing confidence intervals around generalizability coefficients is a challenging task in generalizability theory, with small sample sizes adding to the difficulty. This research compared four LMM-based methods and found that a Bootstrap technique called parametric methods with spherical random effects consistently produced more accurate confidence intervals. Recommendations were made for including confidence intervals with point estimates when computing generalizability coefficients.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2022)
Article
Psychology, Educational
Ren Liu, Haiyan Liu, Dexin Shi, Zhehan Jiang
Summary: This article discusses the use of assessments with small, similar, or repetitive tasks in educational, neurocognitive, and psychological contexts. It extends the Rasch Poisson counts model (RPCM) into diagnostic classification models (DCMs) by applying a Poisson distribution. A framework of Poisson DCMs is proposed and demonstrated using an operational dataset, with recommendations for future research provided.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2022)
Article
Mathematics, Interdisciplinary Applications
Carmen Ximenez, Alberto Maydeu-Olivares, Dexin Shi, Javier Revuelta
Summary: The behavior of fit indices in CFA models depends on factors such as model size, sample size, and measurement quality. Biased estimators of fit indices result in unstable behavior with changing sample size, making it difficult to establish cutoff values. Unbiased estimators, on the other hand, match the behavior of population parameters and depend on the average R-2 of observed variables and model size.
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL
(2022)
Article
Psychology, Educational
Yuting Han, Jihong Zhang, Zhehan Jiang, Dexin Shi
Summary: This article explores the behavior of AUC in fitting IRT models and investigates its appropriateness through simulation experiments under different conditions. The results show that AUC has advantages in certain cases when fitting IRT models, but disadvantages are obvious when the true model is unidimensional.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2023)
Article
Mathematics, Interdisciplinary Applications
Raul Correa Ferraz, Alberto Maydeu-Olivares, Dexin Shi
Summary: Previous research has shown that bootstrapped p-values for the chi-square test of model fit are accurate for small models but not as accurate for small sample sizes. As the number of variables increases, bootstrapped p-values may become too conservative and less accurate. Therefore, they may not be recommended for assessing model fit.
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL
(2022)
Article
Psychology, Educational
Xu Jiang, Netanel Paley, Dexin Shi
Summary: This study analyzed the psychometric properties of the ITEB-Q questionnaire and found good factor structure and measurement invariance in a sample of adolescents, with female adolescents scoring lower on emotion and thought than male counterparts, and Black adolescents scoring higher across the three subscales than White adolescents.
Article
Psychology, Educational
Zhehan Jiang, Yuting Han, Lingling Xu, Dexin Shi, Ren Liu, Jinying Ouyang, Fen Cai
Summary: This study proposes a machine learning-based imputation technique called chaining random forests (CRF) for equating tasks within the nonequivalent groups with anchor test (NEAT) design. Through a simulation study, the equating performance of various methods, including traditional equating methods and CRF-based methods, is compared. The findings indicate that CRF-based methods yield more accurate equated scores in short-length tests with small samples.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2023)
Editorial Material
Mathematics, Interdisciplinary Applications
Raul C. Ferraz, Alberto Maydeu-Olivares
MULTIVARIATE BEHAVIORAL RESEARCH
(2023)
Article
Mathematics, Interdisciplinary Applications
Yunhang Yin, Dexin Shi, Amanda J. Fairchild
Summary: This study aimed to investigate the effect of model size on the root mean square error of approximation (RMSEA) under nonnormal data, and compared the performance of three methods for computing RMSEA. The results suggested that the normal theory method should not be used under nonnormal data, and researchers should consider using either the BSL or the Lai method. The Lai method is recommended for fitting very large models under nonnormal data.
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL
(2023)
Article
Psychology, Educational
Dexin Shi, Bo Zhang, Ren Liu, Zhehan Jiang
Summary: This paper introduces the methods of using multiple imputation (MI) technique to handle missing data in ordinal factor analysis models, and proposes the use of standardized root mean squared residual (SRMR) and root mean square error of approximation (RMSEA) to assess the fit of the models. Simulation results show that the proposed methods can provide accurate point and interval estimates for both SRMR and RMSEA, especially in conditions with larger sample sizes, less missing data, more response categories, and higher degrees of misfit. Implications and recommendations based on the findings are discussed.
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT
(2023)
Article
Mathematics, Interdisciplinary Applications
Dingjing Shi, Dexin Shi, Amanda J. Fairchild
Summary: This study proposes a Bayesian variable selection approach for exploratory mediation analysis, which can select mediators and quantify their posterior probabilities. Monte Carlo simulation studies show that the method has high statistical power in selecting mediating effects and low Type I error rate in excluding null effects. The proposed method offers a unique gain by estimating the probability of a mediating effect through the posterior distribution and quantifying the variable's influence on a continuum scale, which conventional p-value-based mediation methods and the LASSO method lack.
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL
(2023)
Article
Psychology, Clinical
Bethany A. A. Bell, Stephen G. G. Taylor, Alex M. M. Roberts, Dexin Shi, Katelynn Burgess, Caroline Hough, Kate Flory
Summary: The Strengths and Difficulties Questionnaire (SDQ) is commonly used to assess behavioral and emotional symptoms and strengths among children and adolescents. However, its underlying factor structure remains a topic of debate. Using data from a large epidemiologic study in the Southeastern United States, we found support for the original five-factor model of the SDQ when using teacher-reported data from both elementary and secondary school children.