1
|
Kline RB. How to evaluate local fit (residuals) in large structural equation models. INTERNATIONAL JOURNAL OF PSYCHOLOGY 2024. [PMID: 39359027 DOI: 10.1002/ijop.13252] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/19/2023] [Accepted: 09/02/2024] [Indexed: 10/04/2024]
Abstract
Consistent with reporting standards for structural equation modelling (SEM), model fit should be evaluated at two different levels, global and local. Global fit concerns the overall or average correspondence between the entire data matrix and the model, given the parameter estimates for the model. Local fit is evaluated at the level of the residuals, or differences between observed and predicted associations for every pair of measured variables in the model. It can happen that models with apparently satisfactory global fit can nevertheless have problematic local fit. This may be especially true for relatively large models with many variables, where serious misspecification is indicated by some larger residuals, but their contribution to global fit is diluted when averaged together with all the other smaller residuals. It can be challenging to evaluate local fit in large models with dozens or even hundreds of variables and corresponding residuals. Thus, the main goal of this tutorial is to offer suggestions about how to efficiently evaluate and describe local fit for large structural equation models. An empirical example is described where all data, syntax and output files are freely available to readers.
Collapse
Affiliation(s)
- Rex B Kline
- Department of Psychology, Concordia University, Montréal, Canada
| |
Collapse
|
2
|
Rogers P. Best practices for your confirmatory factor analysis: A JASP and lavaan tutorial. Behav Res Methods 2024; 56:6634-6654. [PMID: 38480677 DOI: 10.3758/s13428-024-02375-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 02/17/2024] [Indexed: 08/30/2024]
Abstract
Confirmatory factor analysis (CFA) is a fundamental method for evaluating the internal structural validity of measurement instruments. In most CFA applications, the measurement model serves as a means to an end rather than an end in itself. To select the appropriate model, prior validity evidence is crucial, and items are typically assessed on an ordinal scale, which has been used in the applied social sciences. However, textbooks on structural equation modeling (SEM) often overlook this specific case, focusing on applications estimable using maximum likelihood (ML) instead. Unfortunately, several popular commercial SEM software packages lack suitable solutions for handling this 'typical CFA', leading to confusion and suboptimal decision-making when conducting CFA in this context. This article conceptually contributes to this ongoing discussion by presenting a set of guidelines for conducting a typical CFA, drawing from recent empirical research. We provide a practical contribution by introducing and developing a tutorial example within the JASP and lavaan software platforms. Supplementary materials such as videos, files, and scripts are freely available.
Collapse
Affiliation(s)
- Pablo Rogers
- Federal University of Uberlândia, João Naves de Ávila 2121, 38.408-100, Uberlândia, Minas Gerais, Brazil.
| |
Collapse
|
3
|
Yang L, Tan X, Lang R, Wang T, Li K. Reliability and validity of the Chinese version of the doomscrolling scale and the mediating role of doomscrolling in the bidirectional relationship between insomnia and depression. BMC Psychiatry 2024; 24:565. [PMID: 39160461 PMCID: PMC11334331 DOI: 10.1186/s12888-024-06006-5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 03/25/2024] [Accepted: 08/08/2024] [Indexed: 08/21/2024] Open
Abstract
BACKGROUND Doomscrolling behavior is very common among college students. The purpose of this study was to evaluate the reliability and validity of the Chinese version of the Doomscrolling Scale, thus providing a scientific basis for its application among Chinese university students. METHODS The Chinese version of Doomscrolling Scale was developed through translation and revision of the original scale, conducting item and factor analysis, and validating it with validation factor analysis. The psychometric properties of the Doomscrolling Scale were assessed in 2885 Chinese university students. RESULTS The internal consistency coefficients, two-month test-retest reliability, and split-half reliability of the Chinese version of the Doomscrolling Scale (including the 15-item and the 4-item short version) were high, and the mono-factorial scales fitted well to the theoretical model. Scores on the Chinese version of the Doomscrolling Scale were significantly associated with depression, anxiety, and smartphone addiction. The structural equation model indicates that doomscrolling can mediate the bidirectional relationship between insomnia disorder and depression. CONCLUSIONS The revised Chinese version of the Doomscrolling Scale is valid and reliable, which can facilitate research in this field. The association between doomscrolling and various mental disorders has been confirmed, and further research should be conducted to investigate its mechanisms of action.
Collapse
Affiliation(s)
- Lu Yang
- School of Educational Sciences, Chongqing Normal University, Chongqing, 401331, China
- Chongqing Key Laboratory of Psychological Diagnosis and Education Technology for Children with Special Needs, Chongqing, China
| | - Xuejiao Tan
- Department of Medical English, School of Basic Medicine, Army Medical University, Chongqing, 400038, China
| | - Rui Lang
- Chongqing Vocational Institute of Safety & Technology, Chongqing, China
| | - Tao Wang
- School of Educational Sciences, Chongqing Normal University, Chongqing, 401331, China.
- Chongqing Key Laboratory of Psychological Diagnosis and Education Technology for Children with Special Needs, Chongqing, China.
| | - Kuiliang Li
- Department of Medical English, School of Basic Medicine, Army Medical University, Chongqing, 400038, China.
- School of Psychology, Shaanxi Normal University, Xi'an, China.
| |
Collapse
|
4
|
Carmichael J, Ponsford J, Gould KR, Tiego J, Forbes MK, Kotov R, Fornito A, Spitz G. A Transdiagnostic, Hierarchical Taxonomy of Psychopathology Following Traumatic Brain Injury (HiTOP-TBI). J Neurotrauma 2024. [PMID: 38970424 DOI: 10.1089/neu.2024.0006] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 07/08/2024] Open
Abstract
Psychopathology, including depression, anxiety, and post-traumatic stress, is a significant yet inadequately addressed feature of moderate-severe traumatic brain injury (TBI). Progress in understanding and treating post-TBI psychopathology may be hindered by limitations associated with conventional diagnostic approaches, specifically the Diagnostic and Statistical Manual of Mental Disorders (DSM) and International Classification of Diseases (ICD). The Hierarchical Taxonomy of Psychopathology (HiTOP) offers a promising, transdiagnostic alternative to psychiatric classification that may more effectively capture the experiences of individuals with TBI. However, HiTOP lacks validation in the TBI population. To address this gap, we administered a comprehensive questionnaire battery, including 56 scales assessing homogeneous symptom components and maladaptive traits within HiTOP, to 410 individuals with moderate-severe TBI. We evaluated the reliability and unidimensionality of each scale and revised those with psychometric problems. Using a top-down, exploratory latent variable approach (bass-ackwards modeling), we subsequently constructed a hierarchical model of psychopathological dimensions tailored to TBI. The results showed that, relative to norms, participants with moderate-severe TBI experienced greater problems in the established HiTOP internalizing and detachment spectra, but fewer problems with thought disorder and antagonism. Fourteen of the 56 scales demonstrated psychometric problems, which often appeared reflective of the TBI experience and associated disability. The Hierarchical Taxonomy of Psychopathology Following Traumatic Brain Injury (HiTOP-TBI) model encompassed broad internalizing and externalizing spectra, splitting into seven narrower dimensions: Detachment, Dysregulated Negative Emotionality, Somatic Symptoms, Compensatory and Phobic Reactions, Self-Harm and Psychoticism, Rigid Constraint, and Harmful Substance Use. This study presents the most comprehensive empirical classification of psychopathology after TBI to date. It introduces a novel, TBI-specific transdiagnostic questionnaire battery and model, which addresses the limitations of conventional DSM and ICD diagnoses. The empirical structure of psychopathology after TBI largely aligned with the established HiTOP model (e.g., a detachment spectrum). However, these constructs need to be interpreted in relation to the unique experiences associated with TBI (e.g., considering the injury's impact on the person's social functioning). By overcoming the limitations of conventional diagnostic approaches, the HiTOP-TBI model has the potential to accelerate our understanding of the causes, correlates, consequences, and treatment of psychopathology after TBI.
Collapse
Affiliation(s)
- Jai Carmichael
- Monash-Epworth Rehabilitation Research Centre, School of Psychological Sciences, Monash University, Melbourne, Australia
| | - Jennie Ponsford
- Monash-Epworth Rehabilitation Research Centre, School of Psychological Sciences, Monash University, Melbourne, Australia
| | - Kate Rachel Gould
- Monash-Epworth Rehabilitation Research Centre, School of Psychological Sciences, Monash University, Melbourne, Australia
| | - Jeggan Tiego
- Turner Institute for Brain and Mental Health, School of Psychological Sciences, Monash University, Melbourne, Australia
| | - Miriam K Forbes
- School of Psychological Sciences, Macquarie University, Sydney, Australia
| | - Roman Kotov
- Stony Brook University, New York, New York, USA
| | - Alex Fornito
- Turner Institute for Brain and Mental Health, School of Psychological Sciences, Monash University, Melbourne, Australia
| | - Gershon Spitz
- Monash-Epworth Rehabilitation Research Centre, School of Psychological Sciences, Monash University, Melbourne, Australia
- Department of Neuroscience, Central Clinical School, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Australia
| |
Collapse
|
5
|
Groskurth K, Bluemke M, Lechner CM. Why we need to abandon fixed cutoffs for goodness-of-fit indices: An extensive simulation and possible solutions. Behav Res Methods 2024; 56:3891-3914. [PMID: 37640961 PMCID: PMC11133148 DOI: 10.3758/s13428-023-02193-3] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 07/07/2023] [Indexed: 08/31/2023]
Abstract
To evaluate model fit in confirmatory factor analysis, researchers compare goodness-of-fit indices (GOFs) against fixed cutoff values (e.g., CFI > .950) derived from simulation studies. Methodologists have cautioned that cutoffs for GOFs are only valid for settings similar to the simulation scenarios from which cutoffs originated. Despite these warnings, fixed cutoffs for popular GOFs (i.e., χ2, χ2/df, CFI, RMSEA, SRMR) continue to be widely used in applied research. We (1) argue that the practice of using fixed cutoffs needs to be abandoned and (2) review time-honored and emerging alternatives to fixed cutoffs. We first present the most in-depth simulation study to date on the sensitivity of GOFs to model misspecification (i.e., misspecified factor dimensionality and unmodeled cross-loadings) and their susceptibility to further data and analysis characteristics (i.e., estimator, number of indicators, number and distribution of response options, loading magnitude, sample size, and factor correlation). We included all characteristics identified as influential in previous studies. Our simulation enabled us to replicate well-known influences on GOFs and establish hitherto unknown or underappreciated ones. In particular, the magnitude of the factor correlation turned out to moderate the effects of several characteristics on GOFs. Second, to address these problems, we discuss several strategies for assessing model fit that take the dependency of GOFs on the modeling context into account. We highlight tailored (or "dynamic") cutoffs as a way forward. We provide convenient tables with scenario-specific cutoffs as well as regression formulae to predict cutoffs tailored to the empirical setting of interest.
Collapse
Affiliation(s)
- Katharina Groskurth
- GESIS - Leibniz Institute for the Social Sciences, Mannheim, Germany.
- University of Mannheim, Graduate School of Economic and Social Sciences, Mannheim, Germany.
| | - Matthias Bluemke
- GESIS - Leibniz Institute for the Social Sciences, Mannheim, Germany
- Technical University of Darmstadt, Darmstadt, Germany
| | - Clemens M Lechner
- GESIS - Leibniz Institute for the Social Sciences, Mannheim, Germany
| |
Collapse
|
6
|
Black L, Humphrey N, Panayiotou M, Marquez J. Mental Health and Well-being Measures for Mean Comparison and Screening in Adolescents: An Assessment of Unidimensionality and Sex and Age Measurement Invariance. Assessment 2024; 31:219-236. [PMID: 36864693 PMCID: PMC10822075 DOI: 10.1177/10731911231158623] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 03/04/2023]
Abstract
Adolescence is a period of increased vulnerability for low well-being and mental health problems, particularly for girls and older adolescents. Accurate measurement via brief self-report is therefore vital to understanding prevalence, group trends, screening efforts, and response to intervention. We drew on data from the #BeeWell study (N = 37,149, aged 12-15) to consider whether sum-scoring, mean comparisons, and deployment for screening were likely to show bias for eight such measures. Evidence for unidimensionality, considering dynamic fit confirmatory factor models, exploratory graph analysis, and bifactor modeling, was found for five measures. Of these five, most showed a degree of non-invariance across sex and age likely incompatible with mean comparison. Effects on selection were minimal, except sensitivity was substantially lower in boys for the internalizing symptoms measure. Measure-specific insights are discussed, as are general issues highlighted by our analysis, such as item reversals and measurement invariance.
Collapse
|
7
|
Yin B, Shen Y. Development and Validation of the Compensatory Belief Scale for the Internet Instant Gratification Behavior. Heliyon 2024; 10:e23972. [PMID: 38268608 PMCID: PMC10805770 DOI: 10.1016/j.heliyon.2024.e23972] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/10/2023] [Revised: 11/06/2023] [Accepted: 01/01/2024] [Indexed: 01/26/2024] Open
Abstract
Objective The concept of "compensatory belief" pertains to an individual's conviction that the negative consequences of a specific behavior can be counteracted by engaging in a subsequent positive behavior. This study sought to devise a scale tailored to gauge compensatory beliefs concerning internet instant gratification behaviors. Methods Utilizing the Credamo online questionnaire platform, data were amassed from two distinctive cohorts: organizational employees and college students. A collective 1064 responses were amassed. Results The newly created Compensatory Belief Scale for Internet Instant Gratification Behavior was bifurcated into two versions: one tailored for employees (CBS-IIGB-E) and the other for college students (CBS-IIGB-S). Through an exploratory factor analysis, two factors were discerned, namely "Compensatory Beliefs for Working/Studying " and "Compensatory Beliefs for Resting". A confirmatory factor analysis validated this two-factor model with the following metrics for the employee version: SBχ2 = 54.88, df = 32, CFI = 0.974, TLI = 0.964, RMSEA = 0.064, SRMR = 0.047, and for the student version: SBχ2 = 19.26,df = 19,CFI = 0.999,TLI = 0.999,RMSEA = 0.008,SRMR = 0.033. The scores on the Internet Addiction Scale and the Smartphone Addiction Scale showed a positive correlation with the overall scores of the CBS-IIGB-E and the scores of its two factors. Conversely, the scores on the Self-control Scale exhibited a negative correlation with the total scores of the CBS-IIGB-E and the scores of its two factors. The correlation pattern with the criterion scales for the CBS-IIGB-S mirrored that of the CBS-IIGB-E, with the exception that the Self-control Scale only correlated with the two factors and not the overall score of the CBS-IIGB-S. The McDonald's Omega coefficients for the two factors of the CBS-IIGB-E were 0.84 and 0.86 respectively, and those for the CBS-IIGB-S were 0.82 and 0.87 respectively. Both scales demonstrated a test-retest reliability of 0.76. Significant differences in the scale scores across diverse target groups were observed in the simulation game of internet instant gratification behavior, thereby validating the predictive validity of the developed scales. Conclusion The Compensatory Belief Scale for Internet Instant Gratification Behavior (CBS-IIGB) is a reliable and valid tool for measuring compensatory belief in situations where the allure of immediate internet gratification comes into conflict with long-term objectives, among both organizational employees (CBS-IIGB-E) and college students (CBS-IIGB-S).
Collapse
Affiliation(s)
- Bin Yin
- Laboratory for Learning and Behavioral Sciences, School of Psychology, Fujian Normal University, Fuzhou, 350117, Fujian, China
- Department of Applied Psychology, School of Psychology, Fujian Normal University, 350117, Fujian, China
| | - Yong Shen
- Laboratory for Learning and Behavioral Sciences, School of Psychology, Fujian Normal University, Fuzhou, 350117, Fujian, China
| |
Collapse
|
8
|
Jahrami H, Saif Z, Chen W, Helmy M, Ghazzawi H, Trabelsi K, Natan Pires G, Bragazzi NL, Pandi-Perumal SR, Seeman MV. Development and validation of a questionnaire (GHOST) to assess sudden, unexplained communication exclusion or "ghosting". Heliyon 2023; 9:e17066. [PMID: 37484221 PMCID: PMC10361225 DOI: 10.1016/j.heliyon.2023.e17066] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/19/2022] [Revised: 06/05/2023] [Accepted: 06/06/2023] [Indexed: 07/25/2023] Open
Abstract
The topic of "ghosting" as a method of terminating a relationship has been discussed in both popular media and academic circles. Although research on this issue is scarce, the concept has acquired popularity and gained scholarly attention. A reliable and valid measure of this phenomenon does not, however, exist. GHOST (The Ghosting Questionnaire) was designed and psychometrically tested to explore ghostee experiences. A total of 811 adults participated in an online survey to test this instrument. It was developed based on a thorough analysis of research on the topic of ghosting using a phenomenological qualitative method to identify ghosting domains and generate questionnaire items. In terms of content validity and construct validity, the final version of the measure was found to be satisfactory. GHOST was found to have adequate internal consistency - scores of 0.74, 0.74, and 0.80, indicating acceptable Cronbach's alpha, McDonald's omega, and ordinal's alpha coefficients, respectively. Factor analyses found the GHOST questionnaire to be a valid and reliable instrument that can be used for screening ghosting experiences and for designing community-based distress prevention and intervention programs. A dynamic fit index (DFI) cutoffs approach was also used and showed robust fitting.
Collapse
Affiliation(s)
- Haitham Jahrami
- Government Hospitals, Manama, Bahrain
- Department of Psychiatry, College of Medicine and Medical Sciences, Arabian Gulf University, Manama, Bahrain
| | | | - Wen Chen
- Department of Child Psychology, The Children's Hospital, Zhejiang University School of Medicine, National Clinical Research Center for Child Health, National Children's Regional Medical Center, Hangzhou, Zhejiang, China
| | - Mai Helmy
- Psychology Department, College of Education, Sultan Qaboos University, Muscat, 123, Oman
- Psychology Department, Menoufia University, Shebin El-Kom, 32511, Egypt
| | - Hadeel Ghazzawi
- Nutrition and Food Technology Department, Agriculture School, The University of Jordan, P. O. Box 11942, Jordan
| | - Khaled Trabelsi
- High Institute of Sport and Physical Education of Sfax, University of Sfax, Sfax, 3000, Tunisia
- Research Laboratory: Education, Motricity, Sport and Health, EM2S, LR19JS01, University of Sfax, Sfax, 3000, Tunisia
| | | | - Nicola L. Bragazzi
- Human Nutrition Unit, Department of Food and Drugs, University of Parma Medical School, Parma, Italy
| | - Seithikurippu R. Pandi-Perumal
- Division of Research and Development, Lovely Professional University, Phagwara, Punjab, 144411, India
- Saveetha Medical College and Hospitals, Saveetha Institute of Medical and Technical Sciences, Saveetha University, Chennai, Tamil Nadu, India
| | | |
Collapse
|
9
|
Carmichael J, Spitz G, Gould KR, Johnston L, Samiotis A, Ponsford J. Bifactor analysis of the Hospital Anxiety and Depression Scale (HADS) in individuals with traumatic brain injury. Sci Rep 2023; 13:8017. [PMID: 37198250 PMCID: PMC10192445 DOI: 10.1038/s41598-023-35017-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/27/2023] [Accepted: 05/11/2023] [Indexed: 05/19/2023] Open
Abstract
Anxiety and depression symptoms are commonly experienced after traumatic brain injury (TBI). However, studies validating measures of anxiety and depression for this population are scarce. Using novel indices derived from symmetrical bifactor modeling, we evaluated whether the Hospital Anxiety and Depression Scale (HADS) reliably differentiated anxiety and depression in 874 adults with moderate-severe TBI. The results showed that there was a dominant general distress factor accounting for 84% of the systematic variance in HADS total scores. The specific anxiety and depression factors accounted for little residual variance in the respective subscale scores (12% and 20%, respectively), and overall, minimal bias was found in using the HADS as a unidimensional measure. Further, in a subsample of 184 participants, the HADS subscales did not clearly discriminate between formal anxiety and depressive disorders diagnosed via clinical interview. Results were consistent when accounting for degree of disability, non-English speaking background, and time post-injury. In conclusion, variance in HADS scores after TBI predominately reflects a single underlying latent variable. Clinicians and researchers should exercise caution in interpreting the individual HADS subscales and instead consider using the total score as a more valid, transdiagnostic measure of general distress in individuals with TBI.
Collapse
Affiliation(s)
- Jai Carmichael
- Monash-Epworth Rehabilitation Research Centre, Epworth HealthCare, Melbourne, Australia.
- School of Psychological Sciences, Turner Institute for Brain and Mental Health, Monash University, Clayton, Australia.
| | - Gershon Spitz
- Monash-Epworth Rehabilitation Research Centre, Epworth HealthCare, Melbourne, Australia
- School of Psychological Sciences, Turner Institute for Brain and Mental Health, Monash University, Clayton, Australia
- Department of Neuroscience, Central Clinical School, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Australia
| | - Kate Rachel Gould
- Monash-Epworth Rehabilitation Research Centre, Epworth HealthCare, Melbourne, Australia
- School of Psychological Sciences, Turner Institute for Brain and Mental Health, Monash University, Clayton, Australia
| | - Lisa Johnston
- Monash-Epworth Rehabilitation Research Centre, Epworth HealthCare, Melbourne, Australia
| | - Alexia Samiotis
- Monash-Epworth Rehabilitation Research Centre, Epworth HealthCare, Melbourne, Australia
- School of Psychological Sciences, Turner Institute for Brain and Mental Health, Monash University, Clayton, Australia
| | - Jennie Ponsford
- Monash-Epworth Rehabilitation Research Centre, Epworth HealthCare, Melbourne, Australia
- School of Psychological Sciences, Turner Institute for Brain and Mental Health, Monash University, Clayton, Australia
| |
Collapse
|
10
|
McNeish D. Generalizability of Dynamic Fit Index, Equivalence Testing, and Hu & Bentler Cutoffs for Evaluating Fit in Factor Analysis. MULTIVARIATE BEHAVIORAL RESEARCH 2023; 58:195-219. [PMID: 36787523 DOI: 10.1080/00273171.2022.2163477] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/18/2023]
Abstract
Factor analysis is often used to model scales created to measure latent constructs, and internal structure validity evidence is commonly assessed with indices like RMSEA, and CFI. These indices are essentially effect size measures and definitive benchmarks regarding which values connote reasonable fit have been elusive. Simulations from the 1990s suggesting possible benchmark values are among the most highly cited methodological papers across any discipline. However, simulations have suggested that fixed benchmarks do not generalize well - fit indices are systematically impacted by characteristics like the number of items and the magnitude of the loadings, so fixed benchmarks can confound misfit with model characteristics. Alternative frameworks for creating customized, model-specific benchmarks have recently been proposed to circumvent these issues but they have not been systematically evaluated. Motivated by two empirical applications where different methods yield inconsistent conclusions, two simulation studies are performed to assess the ability of three different approaches to correctly classify models that are correct or misspecified across different conditions. Results show that dynamic fit indices and equivalence testing both improved upon the traditional Hu & Bentler benchmarks and dynamic fit indices appeared to be least confounded with model characteristics in the conditions studied.
Collapse
|
11
|
Wolf MG, McNeish D. dynamic : An R Package for Deriving Dynamic Fit Index Cutoffs for Factor Analysis. MULTIVARIATE BEHAVIORAL RESEARCH 2023; 58:189-194. [PMID: 36787513 DOI: 10.1080/00273171.2022.2163476] [Citation(s) in RCA: 12] [Impact Index Per Article: 12.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/18/2023]
Abstract
To evaluate the fit of a confirmatory factor analysis model, researchers often rely on fit indices such as SRMR, RMSEA, and CFI. These indices are frequently compared to benchmark values of .08, .06, and .96, respectively, established by Hu and Bentler (Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1-55). However, these indices are affected by model characteristics and their sensitivity to misfit can change across models. Decisions about model fit can therefore be improved by tailoring cutoffs to each model. The methodological literature has proposed methods for deriving customized cutoffs, although it can require knowledge of linear algebra and Monte Carlo simulation. Given that many empirical researchers do not have training in these technical areas, empirical studies largely continue to rely on fixed benchmarks even though they are known to generalize poorly and can be poor arbiters of fit. To address this, this paper introduces the R package, dynamic, to make computation of dynamic fit index cutoffs (which are tailored to the user's model) more accessible to empirical researchers. dynamic heavily automatizes this process and only requires a lavaan object to automatically conduct several custom Monte Carlo simulations and output fit index cutoffs designed to be sensitive to misfit with the user's model characteristics.
Collapse
Affiliation(s)
- Melissa G Wolf
- Gevirtz Graduate School of Education, University of California Santa Barbara
| | | |
Collapse
|
12
|
Reder M, Soellner R. Factor Structure of the eHEALS. DIAGNOSTICA 2022. [DOI: 10.1026/0012-1924/a000294] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/04/2022]
Abstract
Abstract. The dimensionality of the eHEALS has been the subject of some controversy. Sample populations and language versions vary widely, as do the employed statistical methods to assess dimensionality. In previous research, we assessed the factor structure in two different samples testing 1 vs. 2 and 1 vs. 2 vs. 3 correlated factors. The objective of this reanalysis was to assess whether the 3-factor model fitted better than the 2- and 1-factor models. We analyzed data from a 2009 cross-sectional survey on health literacy in grade 12 ( n = 327) using CFA. All factor models of the eHEALS showed unsatisfactory model fit. A subsequent exploratory bifactor analysis confirmed multidimensionality and indicated that Item 2 was problematic. When this item was excluded from the correlated factor models, model fit improved, and the 3-factor model showed the best fit. The results in our sample of 12th-grade students offer some support to the German eHEALS having a 3-factor structure similar to the results from our previous research in women aged 50. The replicability of the fit pattern in a different sample and setting was limited by diverging results on Item 2.
Collapse
Affiliation(s)
- Maren Reder
- Institute of Psychology, University of Hildesheim, Germany
| | | |
Collapse
|
13
|
Birknerova Z, Misko D, Ondrijova I, Tomkova A, Cema V, Cigarska BN. Identification of Preferred Representational Sensory System in Neuro-Linguistic Programming. MARKETING AND MANAGEMENT OF INNOVATIONS 2022. [DOI: 10.21272/mmi.2022.3-10] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/29/2022]
Abstract
The paper highlights the scientific debate on the Neurolinguistic programming (NLP) issue. NLP is a collection of approaches, communication tools, techniques, and perspectives that determine how individuals think and communicate through language. NLP is used to recognize and modify patterns of human behavior. The sensory representational system, or the method for recognizing representational systems, which is made up of five main senses, influences this process. Systematization literary sources and approaches to this issue indicate that three sensory representational systems exist in the NLP approach: visual, auditory, and kinesthetic (VAK), and that the individual’s preferred representational sensory system could explain manifested behavior and characteristics in the managerial and marketing sphere. The central purpose of the research and the significance of choice made about this area of interest is to determine each individual’s preferred representational sensory system (VAK) utilizing the original PRSS-VAK methodology. The methodological research tool was the PRSS-VAK methodology which contains nine statements, which are assessed on a scale from 1 (the least describes me) to 4 (the most describes me). The PRSS-VAK methodology would help to comprehend patterns of an individual’s behavior and allied cognitive or emotional processes. EFA (Exploratory Factor Analysis) with Varimax rotation was used to verify the methodology on a sample of 214 respondents from the Slovak Republic, and CFA (Confirmatory Factor Analysis) was used to validate the structure on a sample of 268 respondents from the Slovak Republic. This research empirically and theoretically confirms that one of the preferred representational sensory systems may be dominant. However, this may change regarding the current situation (stimulus, impulse). The research results could be beneficial as a springboard not only for researchers concerning this issue. It also indicates that quantitative research does not determine exactly to which category (visual, auditory, or kinesthetic) a certain individual belongs. Using the identification of a preferred representational sensory system could help to facilitate both management and marketing communications.
Collapse
Affiliation(s)
| | - David Misko
- The University of Presov in Presov, Slovakia
| | | | | | | | | |
Collapse
|