scholarly journals Assessment, Testing and Comparison of Statistical Models using R

2021 ◽  
Author(s):  
Daniel Lüdecke ◽  
Mattan S. Ben-Shachar ◽  
Indrajeet Patil ◽  
Philip Waggoner ◽  
Dominique Makowski
Keyword(s):  
Regression Models ◽  
Model Fit ◽  
Fit Indices ◽  
Diagnostic Plots ◽  
Fit Statistics ◽  
Best Fitting ◽  
Competing Models ◽  
Crucial Part ◽  
Fitting Model ◽  
Consistent Approach

A crucial part of statistical analysis is evaluating a model's quality and fit, or performance. During analysis, especially with regression models, investigating the fit of models to data also often involves selecting the best fitting model amongst many competing models. Upon investigation, fit indices should also be reported both visually and numerically to bring readers in on the investigative effort. While functions to build and produce diagnostic plots or to compute fit statistics exist, these are located across many packages, which results in a lack of a unique and consistent approach to assess the performance of many types of models. The result is a difficult-to-navigate, unorganized ecosystem of individual packages with different syntax, making it onerous for researchers to locate and use fit indices relevant for their unique purposes. The performance package in R fills this gap by offering researchers a suite of intuitive functions with consistent syntax for computing, building, and presenting regression model fit statistics and visualizations.

scholarly journals A comprehensive analysis of the factor structure of the Beck Depression Inventory-II in a sample of outpatients with adjustment disorder and depressive episode

2017 ◽  
Vol 35 (1) ◽  
pp. 53-61 ◽  
Author(s):  
E. McElroy ◽  
P. Casey ◽  
G. Adamson ◽  
P. Filippopoulos ◽  
M. Shevlin
Keyword(s):  
Beck Depression Inventory ◽  
Factor Structure ◽  
Depressive Episode ◽  
Adjustment Disorder ◽  
Scale Scores ◽  
Best Fitting ◽  
Competing Models ◽  
Multidimensional Models ◽  
Unidimensional Scale ◽  
Fitting Model

ObjectivesDespite being commonly used in research and clinical practice, the evidence regarding the factor structure of the Beck Depression Inventory-II (BDI-II) remains equivocal and this has implications on how the scale scores should be aggregated. Researchers continue to debate whether the BDI-II is best viewed as a unidimensional scale, or whether specific subscales have utility. The present study sought to test a comprehensive range of competing factor analytic models of the BDI-II, including traditional non-hierarchical multidimensional models and confirmatory bifactor models.MethodParticipants (n=370) were clinical outpatients diagnosed with either depressive episode or adjustment disorder. Confirmatory factor analysis and confirmatory bifactor modelling were used to test 15 competing models. The unidimensionality of the best fitting model was assessed using three strength indices (explained common variance, percentage of uncontaminated correlations and ω hierarchical).ResultsOverall, bifactor solutions provided superior fit than both unidimensional and non-hierarchical multidimensional models. The best fitting model consisted of a general depression factor and three specific factors: cognitive, somatic and affective. High factor loadings and strength indices for the general depression factor supported the view that the BDI-II measures a single latent construct.ConclusionsThe BDI-II should primarily be viewed as a unidimensional scale, and should be scored as such. Although it is not recommended that scores on individual subscales are used in isolation, they may prove useful in clinical assessment and/or treatment planning if used in conjunction with total scores.

Reliability and validity of the sense of humor scale

2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Daniel Lydon ◽  
Wilson McDermut
Keyword(s):  
Convergent Validity ◽  
Reliability And Validity ◽  
Big Five Personality ◽  
Fit Indices ◽  
Factor Analyses ◽  
Sense Of Humor ◽  
Best Fitting ◽  
Additional Support ◽  
Test Retest Reliability ◽  
Fitting Model

Abstract This study examined the reliability, validity, and factor structure of the sense of humor scale (SHS; McGhee, Paul E. 1999. Health, healing and the amuse system: Humor as survival training, 3rd edn. Dubuque: Kendall/Hunt), a 24-item questionnaire developed to measure overall sense of humor. Participants included 99 adults, 105 undergraduates, and 111 comedians. One-week test-retest reliability was good (r = 0.75). Internal consistency of the overall scale was excellent, and acceptable-to-excellent for the six subscales. Item-total correlations were generally strong. Comedians scored higher than undergraduates and adults, supporting the construct validity of the SHS. Convergent validity was strong as the SHS was positively correlated with the Humor Styles Questionnaire total and its subscales. Our analyses of SHS’s associations with the Big Five personality dimensions led to findings that are consistent with prior research, as the SHS was positively correlated with extraversion and openness to experience, but uncorrelated with neuroticism, agreeableness, and conscientiousness. Factor analyses found a bifactor model to be the best fitting model for the SHS. Ancillary bifactor fit indices provided additional support for the notion that the SHS may not be best described as unidimensional. Thus, it can be argued that the subscales are relevant for both research and applied work as they offer unique contributions.

scholarly journals Modeling and Forecasting of Covid-19 Growth Curve in India

2020 ◽  
Author(s):  
Vikas Kumar Sharma ◽  
Unnati Nigam
Keyword(s):  
Akaike Information Criterion ◽  
Regression Models ◽  
Arima Model ◽  
Information Criterion ◽  
Exponential Smoothing ◽  
Modeling And Forecasting ◽  
Best Fitting ◽  
Auto Regressive ◽  
Fitting Model ◽  
Optimal Change

AbstractIn this article, we analyze the growth pattern of Covid-19 pandemic in India from March 4th to May 15th using regression analysis (exponential and polynomial), auto-regressive integrated moving averages (ARIMA) model as well as exponential smoothing and Holt-Winters models. We found that the growth of Covid-19 cases follows a power regime of (t2, t,..) after the exponential growth. We found the optimal change points from where the Covid-19 cases shift their course of growth from exponential to quadratic and then from quadratic to linear. We have also found the best fitted regression models using the various criteria such as significant p-values, coefficients of determination and ANOVA etc. Further, we search the best fitting ARIMA model for the data using the AIC (Akaike Information Criterion) and CAIC (Consistent Akaike Information Criterion) and provide the forecast of Covid-19 cases for future days. We also use usual exponential smoothing and Holt-Winters models for forecasting purpose. We further found that the ARIMA (2,2,0) model is the best-fitting model for Covid-19 cases in India.

scholarly journals Models for Analyzing Zero-Inflated and Overdispersed Count Data: An Application to Cigarette and Marijuana Use

2018 ◽  
Vol 22 (8) ◽  
pp. 1390-1398 ◽  
Author(s):  
Brian Pittman ◽  
Eugenia Buta ◽  
Suchitra Krishnan-Sarin ◽  
Stephanie S O’Malley ◽  
Thomas Liss ◽  
...  
Keyword(s):  
Count Data ◽  
Regression Models ◽  
Goodness Of Fit ◽  
Negative Binomial ◽  
Research Question ◽  
Model Fit ◽  
Cigarette Use ◽  
Zero Inflation ◽  
Fit Statistics ◽  
Spurious Results

Abstract Introduction This article describes different methods for analyzing counts and illustrates their use on cigarette and marijuana smoking data. Methods The Poisson, zero-inflated Poisson (ZIP), hurdle Poisson (HUP), negative binomial (NB), zero-inflated negative binomial (ZINB), and hurdle negative binomial (HUNB) regression models are considered. The different approaches are evaluated in terms of the ability to take into account zero-inflation (extra zeroes) and overdispersion (variance larger than expected) in count outcomes, with emphasis placed on model fit, interpretation, and choosing an appropriate model given the nature of the data. The illustrative data example focuses on cigarette and marijuana smoking reports from a study on smoking habits among youth e-cigarette users with gender, age, and e-cigarette use included as predictors. Results Of the 69 subjects available for analysis, 36% and 64% reported smoking no cigarettes and no marijuana, respectively, suggesting both outcomes might be zero-inflated. Both outcomes were also overdispersed with large positive skew. The ZINB and HUNB models fit the cigarette counts best. According to goodness-of-fit statistics, the NB, HUNB, and ZINB models fit the marijuana data well, but the ZINB provided better interpretation. Conclusion In the absence of zero-inflation, the NB model fits smoking data well, which is typically overdispersed. In the presence of zero-inflation, the ZINB or HUNB model is recommended to account for additional heterogeneity. In addition to model fit and interpretability, choosing between a zero-inflated or hurdle model should ultimately depend on the assumptions regarding the zeros, study design, and the research question being asked. Implications Count outcomes are frequent in tobacco research and often have many zeros and exhibit large variance and skew. Analyzing such data based on methods requiring a normally distributed outcome are inappropriate and will likely produce spurious results. This study compares and contrasts appropriate methods for analyzing count data, specifically those with an over-abundance of zeros, and illustrates their use on cigarette and marijuana smoking data. Recommendations are provided.

scholarly journals Perfectionism and Burnout in Junior Athletes: A Three-Month Longitudinal Study

2015 ◽  
Vol 37 (3) ◽  
pp. 305-315 ◽  
Author(s):  
Daniel J. Madigan ◽  
Joachim Stoeber ◽  
Louis Passfield
Keyword(s):  
Risk Factor ◽  
Structural Equation ◽  
Protective Factor ◽  
Equation Modeling ◽  
Junior Athletes ◽  
Athlete Burnout ◽  
Best Fitting ◽  
Competing Models ◽  
Active Training ◽  
Fitting Model

Perfectionism in sports has been shown to be associated with burnout in athletes. Whether perfectionism predicts longitudinal changes in athlete burnout, however, is still unclear. Using a two-wave cross-lagged panel design, the current study examined perfectionistic strivings, perfectionistic concerns, and athlete burnout in 101 junior athletes (mean age 17.7 years) over 3 months of active training. When structural equation modeling was employed to test a series of competing models, the best-fitting model showed opposite patterns for perfectionistic strivings and perfectionistic concerns. Whereas perfectionistic concerns predicted increases in athlete burnout over the 3 mon ths, perfectionistic strivings predicted decreases. The present findings suggest that perfectionistic concerns are a risk factor for junior athletes contributing to the development of athlete burnout whereas perfectionistic strivings appear to be a protective factor.

A Bifactor Model of Trait Trust, Distrust, and Suspicion

2021 ◽  
Author(s):  
Gene M. Alarcon ◽  
August Capiola ◽  
Sarah A. Jessup ◽  
Tyler J. Ryan ◽  
Anthony M. Gibson
Keyword(s):  
Item Response Theory ◽  
Item Response ◽  
General Factor ◽  
Best Fitting ◽  
Competing Models ◽  
Latent Traits ◽  
Relationship Of ◽  
Dispositional Trust ◽  
The Relationship ◽  
Fitting Model

Abstract. We explored competing models using bifactor item response theory (IRT) analyses to determine the relationship between trait measures of trust, distrust, and suspicion. The model with a general factor for all three scales fits the data best. We explored the relationship of the emergent general factor by correlating it with two latent traits: Agreeableness and the Trust facet of Agreeableness. The exploratory findings showed evidence that the general factor from the best-fitting model was practically identical to the Trust facet of Agreeableness. We concluded that trait trust, distrust, and suspicion reside on a continuum represented by the general factor, which is dispositional trust.

A Bootstrap Procedure for Evaluating Goodness-of-Fit Indices of Structural Equation and Confirmatory Factor Models

1989 ◽  
Vol 26 (1) ◽  
pp. 105-111 ◽  
Author(s):  
Paula Fitzgerald Bone ◽  
Subhash Sharma ◽  
Terence A. Shimp
Keyword(s):  
Structural Equation ◽  
Goodness Of Fit ◽  
Factor Models ◽  
Model Fit ◽  
Fit Indices ◽  
Bootstrap Sampling ◽  
Bootstrap Procedure ◽  
Number Of Factors ◽  
Confirmatory Factor ◽  
Competing Models

The authors propose a bootstrap procedure for evaluating the goodness-of-fit indices for structural equation and confirmatory factor models. Monté Carlo simulations are applied to obtain a bootstrap sampling distribution (BSD) for each fit statistic. Then the BSD is used to evaluate model fit. Because the BSD takes into consideration sample size and model characteristics (e.g., number of factors, number of indicators per factor), its application in the proposed procedure makes it possible to compare the fits of competing models. Two previous studies are reanalyzed in illustrating how to implement the proposed procedure.

scholarly journals Assessing Psychometric Properties of the Italian Version of the Heartland Forgiveness Scale

2020 ◽  
Vol 11 ◽  
Author(s):  
Simone Consoli ◽  
Alessandro Rossi ◽  
Laura Y. Thompson ◽  
Clarissa Volpi ◽  
Stefania Mannarini ◽  
...  
Keyword(s):  
Scientific Literature ◽  
Model Fit ◽  
Fit Indices ◽  
Factorial Structure ◽  
Italian Version ◽  
Good Reliability ◽  
Dispositional Forgiveness ◽  
Final Sample ◽  
Confirmatory Factor ◽  
Competing Models

Despite increasing popularity and intensive worldwide use, few studies have assessed the validity and factorial structure of the Heartland Forgiveness Scale (HFS). However, scientific literature showed that the original factorial structure of the HFS was not fully replicated and—in addition—the Italian translation is still lacking. To fill this gap, this study aims to extend evidence about the original HFS factorial validity by analyzing the Italian version. The final sample was composed of 523 randomly enrolled participants [139 males (26.6%), 384 females (73.4%)] aged from 18 to 82 years (mean = 42.53, SD = 16.41) who completed the Italian version of the HFS. The confirmatory factor analysis showed good fit indices for the original hierarchical factor solution and a significant decrease in model fit was found for all of the competing models. Also, the Italian version of the HFS revealed good reliability and very good psychometrical properties. Findings suggest that the Italian version of the HFS can be considered a reliable and good psychometrically based instrument for the assessment of dispositional forgiveness of the Self, Other, and Situation.

Forced Zero Cross-Loading Misspecifications in Measurement Component of Structural Equation Models

Methodology ◽  
2014 ◽  
Vol 10 (4) ◽  
pp. 138-152 ◽  
Author(s):  
Hsien-Yuan Hsu ◽  
Susan Troncoso Skidmore ◽  
Yan Li ◽  
Bruce Thompson
Keyword(s):  
Structural Equation ◽  
Structural Model ◽  
Structural Equation Models ◽  
Equation Model ◽  
Model Fit ◽  
Model Specification ◽  
Simulation Design ◽  
Measurement Models ◽  
Fit Statistics ◽  
Path Coefficients

The purpose of the present paper was to evaluate the effect of constraining near-zero parameter cross-loadings to zero in the measurement component of a structural equation model. A Monte Carlo 3 × 5 × 2 simulation design was conducted (i.e., sample sizes of 200, 600, and 1,000; parameter cross-loadings of 0.07, 0.10, 0.13, 0.16, and 0.19 misspecified to be zero; and parameter path coefficients in the structural model of either 0.50 or 0.70). Results indicated that factor pattern coefficients and factor covariances were overestimated in measurement models when near-zero parameter cross-loadings constrained to zero were higher than 0.13 in the population. Moreover, the path coefficients between factors were misestimated when the near-zero parameter cross-loadings constrained to zero were noteworthy. Our results add to the literature detailing the importance of testing individual model specification decisions, and not simply evaluating omnibus model fit statistics.

Is That Measure Really One-Dimensional?

Methodology ◽  
2018 ◽  
Vol 14 (4) ◽  
pp. 188-196 ◽  
Author(s):  
Esther T. Beierl ◽  
Markus Bühner ◽  
Moritz Heene
Keyword(s):  
Model Fit ◽  
Nuisance Parameters ◽  
Factorial Validity ◽  
Fit Indices ◽  
Analysis Model ◽  
Factor Analysis Model ◽  
One Dimensional ◽  
Confirmatory Factor Analysis Model ◽  
Cutoff Values ◽  
Confirmatory Factor

Abstract. Factorial validity is often assessed using confirmatory factor analysis. Model fit is commonly evaluated using the cutoff values for the fit indices proposed by Hu and Bentler (1999) . There is a body of research showing that those cutoff values cannot be generalized. Model fit does not only depend on the severity of misspecification, but also on nuisance parameters, which are independent of the misspecification. Using a simulation study, we demonstrate their influence on measures of model fit. We specified a severe misspecification, omitting a second factor, which signifies factorial invalidity. Measures of model fit showed only small misfit because nuisance parameters, magnitude of factor loadings and a balanced/imbalanced number of indicators per factor, also influenced the degree of misfit. Drawing from our results, we discuss challenges in the assessment of factorial validity.

Sign in / Sign up

Export Citation Format

Share Document