Pub Date : 2010-01-20DOI: 10.1027/1614-2241/A000005
S. Vansteelandt, J. Carpenter, M. Kenward
This article reviews inverse probability weighting methods and doubly robust estimation methods for the analysis of incomplete data sets. We first consider methods for estimating a population mean when the outcome is missing at random, in the sense that measured covariates can explain whether or not the outcome is observed. We then sketch the rationale of these methods and elaborate on their usefulness in the presence of influential inverse weights. We finally outline how to apply these methods in a variety of settings, such as for fitting regression models with incomplete outcomes or covariates, emphasizing the use of standard software programs.
{"title":"Analysis of Incomplete Data Using Inverse Probability Weighting and Doubly Robust Estimators","authors":"S. Vansteelandt, J. Carpenter, M. Kenward","doi":"10.1027/1614-2241/A000005","DOIUrl":"https://doi.org/10.1027/1614-2241/A000005","url":null,"abstract":"This article reviews inverse probability weighting methods and doubly robust estimation methods for the analysis of incomplete data sets. We first consider methods for estimating a population mean when the outcome is missing at random, in the sense that measured covariates can explain whether or not the outcome is observed. We then sketch the rationale of these methods and elaborate on their usefulness in the presence of influential inverse weights. We finally outline how to apply these methods in a variety of settings, such as for fitting regression models with incomplete outcomes or covariates, emphasizing the use of standard software programs.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"6 1","pages":"37-48"},"PeriodicalIF":3.1,"publicationDate":"2010-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292700","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-11-09DOI: 10.1027/1614-2241.5.4.123
R. Bersabé, Teresa Rivas, C. Berrocal
From the proportional odds (PO) model, we obtain general equations to compute multiple cut scores on a test score. This analytical procedure is based on the relationship between a test score (X) and an ordinal outcome variable (Y) with more than two categories. Cut scores are established at the test scores corresponding to the intersection of adjacent category distributions. The application of this procedure is illustrated by an example with data from an actual study on eating disorders (EDs). In this example, two cut scores on the Eating Attitudes Test (EAT-26) are established in order to differentiate between three ordered categories: (1) asymptomatic, (2) symptomatic, and (3) eating disorder. Diagnoses were made from the responses to a self-report (Q-EDD) that operationalizes DSM-IV criteria for EDs. Alternatives to the PO model, when the PO assumption is rejected, are discussed.
{"title":"Obtaining Equations From the Proportional Odds Model to Set Multiple Cut Scores on a Test","authors":"R. Bersabé, Teresa Rivas, C. Berrocal","doi":"10.1027/1614-2241.5.4.123","DOIUrl":"https://doi.org/10.1027/1614-2241.5.4.123","url":null,"abstract":"From the proportional odds (PO) model, we obtain general equations to compute multiple cut scores on a test score. This analytical procedure is based on the relationship between a test score (X) and an ordinal outcome variable (Y) with more than two categories. Cut scores are established at the test scores corresponding to the intersection of adjacent category distributions. The application of this procedure is illustrated by an example with data from an actual study on eating disorders (EDs). In this example, two cut scores on the Eating Attitudes Test (EAT-26) are established in order to differentiate between three ordered categories: (1) asymptomatic, (2) symptomatic, and (3) eating disorder. Diagnoses were made from the responses to a self-report (Q-EDD) that operationalizes DSM-IV criteria for EDs. Alternatives to the PO model, when the PO assumption is rejected, are discussed.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"5 1","pages":"123-130"},"PeriodicalIF":3.1,"publicationDate":"2009-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1027/1614-2241.5.4.123","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292576","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-07-23DOI: 10.1027/1614-2241.5.3.71
M. Eid, Fridtjof W. Nussbeck
Fifty years ago, in 1959, Campbell and Fiske published one of the most influential papers in psychology. In their article Convergent and discriminant validation by the multitraitmultimethod matrix, they argued that it is not sufficient to consider one single operationalization of one construct for purposes of test validation but that multiple measures of multiple constructs are necessary. Campbell and Fiske recommended using at least two methods that are as different as possible for measuring the constructs. Moreover, Campbell and Fiske made clear that it is not possible to get a measure of a trait that is free of method-specific influences. Whenever, in science, we measure a construct (a trait) we have to use a specific measurement method. Therefore, it is the trait and the method that influence the observed score simultaneously. In order to separate methodfrom traitspecific influences, it is thus always necessary to consider more than one trait and more than one method in the validation process. Campbell and Fiske proposed the multitraitmultimethod (MTMM) matrix for analyzing the convergent and discriminant validity. The MTMM matrix consists of the correlations between all multiple measures representing the different traits measured by the different methods. These correlations can be evaluated by several criteria that have been developed by Campbell and Fiske. If the different measures of the same construct are highly correlated, this proves convergent validity. If the different measures of one construct are not correlated with the measures of another construct, this indicates discriminant validity. Campbell and Fiske’s article had and has an enormous influence on psychology (Eid & Diener, 2006). It is the most often cited paper that has ever been published in Psychological Bulletin (Sternberg, 1992). To date, it has been cited 4,735 times (Social Science Citation Index, February 27, 2009, 3:41 pm), and its citation rate is increasing. Their article does not only have an important impact on test validation studies but also has a strong impact on methodological research as many researchers have developed new approaches for analyzing MTMM data and tried to overcome some of the problems and limitations that are related to former approaches of analyzing MTMM matrices. This special issue is dedicated to honoring Campbell and Fiske’s influential work. It presents three different modern approaches for analyzing MTMM data. All contributors use the same data set illustrating their approaches. This enables readers to concentrate on the comparison of the different approaches with respect to the way convergent and discriminant validity can be analyzed as well as how traitand method-specific influences can be identified and quantified. The data consists of three personality traits (extraversion, neuroticism, and conscientiousness) assessed by three raters (one selfand two peer raters). Each scale consists of four items (adjectives such as talkative, conscie
{"title":"The Multitrait-Multimethod Matrix at 50!","authors":"M. Eid, Fridtjof W. Nussbeck","doi":"10.1027/1614-2241.5.3.71","DOIUrl":"https://doi.org/10.1027/1614-2241.5.3.71","url":null,"abstract":"Fifty years ago, in 1959, Campbell and Fiske published one of the most influential papers in psychology. In their article Convergent and discriminant validation by the multitraitmultimethod matrix, they argued that it is not sufficient to consider one single operationalization of one construct for purposes of test validation but that multiple measures of multiple constructs are necessary. Campbell and Fiske recommended using at least two methods that are as different as possible for measuring the constructs. Moreover, Campbell and Fiske made clear that it is not possible to get a measure of a trait that is free of method-specific influences. Whenever, in science, we measure a construct (a trait) we have to use a specific measurement method. Therefore, it is the trait and the method that influence the observed score simultaneously. In order to separate methodfrom traitspecific influences, it is thus always necessary to consider more than one trait and more than one method in the validation process. Campbell and Fiske proposed the multitraitmultimethod (MTMM) matrix for analyzing the convergent and discriminant validity. The MTMM matrix consists of the correlations between all multiple measures representing the different traits measured by the different methods. These correlations can be evaluated by several criteria that have been developed by Campbell and Fiske. If the different measures of the same construct are highly correlated, this proves convergent validity. If the different measures of one construct are not correlated with the measures of another construct, this indicates discriminant validity. Campbell and Fiske’s article had and has an enormous influence on psychology (Eid & Diener, 2006). It is the most often cited paper that has ever been published in Psychological Bulletin (Sternberg, 1992). To date, it has been cited 4,735 times (Social Science Citation Index, February 27, 2009, 3:41 pm), and its citation rate is increasing. Their article does not only have an important impact on test validation studies but also has a strong impact on methodological research as many researchers have developed new approaches for analyzing MTMM data and tried to overcome some of the problems and limitations that are related to former approaches of analyzing MTMM matrices. This special issue is dedicated to honoring Campbell and Fiske’s influential work. It presents three different modern approaches for analyzing MTMM data. All contributors use the same data set illustrating their approaches. This enables readers to concentrate on the comparison of the different approaches with respect to the way convergent and discriminant validity can be analyzed as well as how traitand method-specific influences can be identified and quantified. The data consists of three personality traits (extraversion, neuroticism, and conscientiousness) assessed by three raters (one selfand two peer raters). Each scale consists of four items (adjectives such as talkative, conscie","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"18 1","pages":"71-71"},"PeriodicalIF":3.1,"publicationDate":"2009-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292510","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-07-23DOI: 10.1027/1614-2241.5.3.88
Fridtjof W. Nussbeck, M. Eid, C. Geiser, D. Courvoisier, T. Lischetzke
Many psychologists collect multitrait-multimethod (MTMM) data to assess the convergent and discriminant validity of psychological measures. In order to choose the most appropriate model, the types of methods applied have to be considered. It is shown how the combination of interchangeable and structurally different raters can be analyzed with an extension of the correlated trait-correlated method minus one [CTC(M−1)] model. This extension allows for disentangling individual rater biases (unique method effects) from shared rater biases (common method effects). The basic ideas of this model are presented and illustrated by an empirical example.
{"title":"A CTC(M−1) Model for Different Types of Raters","authors":"Fridtjof W. Nussbeck, M. Eid, C. Geiser, D. Courvoisier, T. Lischetzke","doi":"10.1027/1614-2241.5.3.88","DOIUrl":"https://doi.org/10.1027/1614-2241.5.3.88","url":null,"abstract":"Many psychologists collect multitrait-multimethod (MTMM) data to assess the convergent and discriminant validity of psychological measures. In order to choose the most appropriate model, the types of methods applied have to be considered. It is shown how the combination of interchangeable and structurally different raters can be analyzed with an extension of the correlated trait-correlated method minus one [CTC(M−1)] model. This extension allows for disentangling individual rater biases (unique method effects) from shared rater biases (common method effects). The basic ideas of this model are presented and illustrated by an empirical example.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"5 1","pages":"88-98"},"PeriodicalIF":3.1,"publicationDate":"2009-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292564","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-07-23DOI: 10.1027/1614-2241.5.3.78
F. Oort
Multitrait-multimethod (MTMM) data are characterized by three modes: traits, methods, and subjects. Considering subjects as random, and traits and methods as fixed, stochastic three-mode models can be used to analyze MTMM covariance data. Stochastic three-mode models can be written as linear latent variable models with direct product (DP) restrictions on the parameter matrices (Oort, 1999), yielding three-mode factor models (Bentler & Lee, 1979) and composite direct product models (Browne, 1984) as special cases. DP restrictions on factor loadings and factor correlations facilitate interpretation of the results and enable easy evaluation of the validity requirements of MTMM correlations (Campbell & Fiske, 1959). As an illustrative example, a series of stochastic three-mode models has been fitted to data of three personality traits of 482 students, measured with 12 items, through three methods.
{"title":"Three-Mode Models for Multitrait-Multimethod Data","authors":"F. Oort","doi":"10.1027/1614-2241.5.3.78","DOIUrl":"https://doi.org/10.1027/1614-2241.5.3.78","url":null,"abstract":"Multitrait-multimethod (MTMM) data are characterized by three modes: traits, methods, and subjects. Considering subjects as random, and traits and methods as fixed, stochastic three-mode models can be used to analyze MTMM covariance data. Stochastic three-mode models can be written as linear latent variable models with direct product (DP) restrictions on the parameter matrices (Oort, 1999), yielding three-mode factor models (Bentler & Lee, 1979) and composite direct product models (Browne, 1984) as special cases. DP restrictions on factor loadings and factor correlations facilitate interpretation of the results and enable easy evaluation of the validity requirements of MTMM correlations (Campbell & Fiske, 1959). As an illustrative example, a series of stochastic three-mode models has been fitted to data of three personality traits of 482 students, measured with 12 items, through three methods.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"5 1","pages":"78-87"},"PeriodicalIF":3.1,"publicationDate":"2009-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-05-07DOI: 10.1027/1614-2241.4.2.67
E. Korendijk, C. Maas, M. Moerbeek, P. Heijden
Like in ordinary regression models, in multilevel analysis, homoscedasticity of the residual variances is an assumption that is mostly unchecked. However, in experimental research, the residual variance component at level two may differ in the experimental and the control condition, leading to heteroscedastic second level variances. Using a simulation study, the consequences of ignoring second level heteroscedasticity on the estimation of the fixed and random parameters and their standard errors was investigated. It was found that the standard error of the second level variance is underestimated, but that the estimated fixed parameters of the independent variables, the first level variance and their standard errors are mostly unbiased.
{"title":"The Influence of Misspecification of the Heteroscedasticity on Multilevel Regression Parameter and Standard Error Estimates","authors":"E. Korendijk, C. Maas, M. Moerbeek, P. Heijden","doi":"10.1027/1614-2241.4.2.67","DOIUrl":"https://doi.org/10.1027/1614-2241.4.2.67","url":null,"abstract":"Like in ordinary regression models, in multilevel analysis, homoscedasticity of the residual variances is an assumption that is mostly unchecked. However, in experimental research, the residual variance component at level two may differ in the experimental and the control condition, leading to heteroscedastic second level variances. Using a simulation study, the consequences of ignoring second level heteroscedasticity on the estimation of the fixed and random parameters and their standard errors was investigated. It was found that the standard error of the second level variance is underestimated, but that the estimated fixed parameters of the independent variables, the first level variance and their standard errors are mostly unbiased.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"4 1","pages":"67-72"},"PeriodicalIF":3.1,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292453","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-05-07DOI: 10.1027/1614-2241.4.2.51
A. Kelava, H. Moosbrugger, Polina Dimitruk, K. Schermelleh-engel
Multicollinearity complicates the simultaneous estimation of interaction and quadratic effects in structural equation modeling (SEM). So far, approaches developed within the Kenny-Judd (1984) tradition have failed to specify additional and necessary constraints on the measurement error covariances of the nonlinear indicators. Given that the constraints comprise, in part, latent linear predictor correlations, multicollinearity poses a problem for such approaches. Klein and Moosbrugger’s (2000) latent moderated structural equations approach (LMS) approach does not utilize nonlinear indicators and should therefore not be affected by this problem. In the context of a simulation study, we varied predictor correlation and the number of nonlinear effects in order to compare the performance of three approaches developed for the estimation of simultaneous nonlinear effects: Ping’s (1996) two-step approach, a correctly extended Joreskog-Yang (1996) approach, and LMS. Results show that in contrast to the Joreskog-Ya...
{"title":"Multicollinearity and missing constraints: A comparison of three approaches for the analysis of latent nonlinear effects.","authors":"A. Kelava, H. Moosbrugger, Polina Dimitruk, K. Schermelleh-engel","doi":"10.1027/1614-2241.4.2.51","DOIUrl":"https://doi.org/10.1027/1614-2241.4.2.51","url":null,"abstract":"Multicollinearity complicates the simultaneous estimation of interaction and quadratic effects in structural equation modeling (SEM). So far, approaches developed within the Kenny-Judd (1984) tradition have failed to specify additional and necessary constraints on the measurement error covariances of the nonlinear indicators. Given that the constraints comprise, in part, latent linear predictor correlations, multicollinearity poses a problem for such approaches. Klein and Moosbrugger’s (2000) latent moderated structural equations approach (LMS) approach does not utilize nonlinear indicators and should therefore not be affected by this problem. In the context of a simulation study, we varied predictor correlation and the number of nonlinear effects in order to compare the performance of three approaches developed for the estimation of simultaneous nonlinear effects: Ping’s (1996) two-step approach, a correctly extended Joreskog-Yang (1996) approach, and LMS. Results show that in contrast to the Joreskog-Ya...","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"4 1","pages":"51-66"},"PeriodicalIF":3.1,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292420","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-05-07DOI: 10.1027/1614-2241.4.2.73
Luis M. Lozano, E. García-Cueto, J. Muñiz
The Likert-type format is one of the most widely used in all types of scales in the field of social sciences. Nevertheless, there is no definitive agreement on the number of response categories that optimizes the psychometric properties of the scales. The aim of the present work is to determine in a systematic fashion the number of response alternatives that maximizes the fundamental psychometric properties of a scale: reliability and validity. The study is carried out with data simulated using the Monte Carlo method. We simulate responses to 30 items with correlations between them ranging from 0.2 to 0.9. We also manipulate sample size, analyzing four different sizes: 50, 100, 200, and 500 cases. The number of response options employed ranges from two to nine. The results show that as the number of response alternatives increases, both reliability and validity improve. The optimum number of alternatives is between four and seven. With fewer than four alternatives the reliability and validity decrease, an...
{"title":"Effect of the Number of Response Categories on the Reliability and Validity of Rating Scales","authors":"Luis M. Lozano, E. García-Cueto, J. Muñiz","doi":"10.1027/1614-2241.4.2.73","DOIUrl":"https://doi.org/10.1027/1614-2241.4.2.73","url":null,"abstract":"The Likert-type format is one of the most widely used in all types of scales in the field of social sciences. Nevertheless, there is no definitive agreement on the number of response categories that optimizes the psychometric properties of the scales. The aim of the present work is to determine in a systematic fashion the number of response alternatives that maximizes the fundamental psychometric properties of a scale: reliability and validity. The study is carried out with data simulated using the Monte Carlo method. We simulate responses to 30 items with correlations between them ranging from 0.2 to 0.9. We also manipulate sample size, analyzing four different sizes: 50, 100, 200, and 500 cases. The number of response options employed ranges from two to nine. The results show that as the number of response alternatives increases, both reliability and validity improve. The optimum number of alternatives is between four and seven. With fewer than four alternatives the reliability and validity decrease, an...","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"4 1","pages":"73-79"},"PeriodicalIF":3.1,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1027/1614-2241.4.2.73","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292472","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-01-21DOI: 10.1027/1614-2241.4.1.37
J. Rosel, I. Plewis
Abstract. In this paper we review different structural equation models for the analysis of longitudinal data: (a) univariate models of observable variables, (b) multivariate models of observable variables, (c) models with latent variables, (d) models that are unconditioned or conditioned to other variables (depending on the variability of the independent variables: time-varying or time-invariant, and depending on the type of independent variables: of latent variables or of observable variables), (e) models with interaction of variables, (f) models with nonlinear variables, (g) models with a constant, (h) with single level and multilevel measurement, and (i) other advances in SEM of longitudinal data (latent growth curve model, latent difference score, etc.). We pay more attention to the interaction of variables and to nonlinear transformations of variables because they are not frequently used in empirical investigation. They do, however, offer interesting possibilities to researchers who wish to verify re...
{"title":"Longitudinal Data Analysis with Structural Equations","authors":"J. Rosel, I. Plewis","doi":"10.1027/1614-2241.4.1.37","DOIUrl":"https://doi.org/10.1027/1614-2241.4.1.37","url":null,"abstract":"Abstract. In this paper we review different structural equation models for the analysis of longitudinal data: (a) univariate models of observable variables, (b) multivariate models of observable variables, (c) models with latent variables, (d) models that are unconditioned or conditioned to other variables (depending on the variability of the independent variables: time-varying or time-invariant, and depending on the type of independent variables: of latent variables or of observable variables), (e) models with interaction of variables, (f) models with nonlinear variables, (g) models with a constant, (h) with single level and multilevel measurement, and (i) other advances in SEM of longitudinal data (latent growth curve model, latent difference score, etc.). We pay more attention to the interaction of variables and to nonlinear transformations of variables because they are not frequently used in empirical investigation. They do, however, offer interesting possibilities to researchers who wish to verify re...","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"4 1","pages":"37-50"},"PeriodicalIF":3.1,"publicationDate":"2008-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1027/1614-2241.4.1.37","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2007-09-27DOI: 10.1027/1614-2241.3.3.100
Polina Dimitruk, K. Schermelleh-engel, A. Kelava, H. Moosbrugger
Abstract. Challenges in evaluating nonlinear effects in multiple regression analyses include reliability, validity, multicollinearity, and dichotomization of continuous variables. While reliability and validity issues are solved by employing nonlinear structural equation modeling, multicollinearity remains a problem which may even be aggravated when using latent variable approaches. Further challenges of nonlinear latent analyses comprise the distribution of latent product terms, a problem especially relevant for approaches using maximum likelihood estimation methods based on multivariate normally distributed variables, and unbiased estimates of nonlinear effects under multicollinearity. The only methods that explicitly take the nonnormality of nonlinear latent models into account are latent moderated structural equations (LMS) and quasi-maximum likelihood (QML). In a small simulation study both methods yielded unbiased parameter estimates and correct estimates of standard errors for inferential statistic...
{"title":"Challenges in Nonlinear Structural Equation Modeling","authors":"Polina Dimitruk, K. Schermelleh-engel, A. Kelava, H. Moosbrugger","doi":"10.1027/1614-2241.3.3.100","DOIUrl":"https://doi.org/10.1027/1614-2241.3.3.100","url":null,"abstract":"Abstract. Challenges in evaluating nonlinear effects in multiple regression analyses include reliability, validity, multicollinearity, and dichotomization of continuous variables. While reliability and validity issues are solved by employing nonlinear structural equation modeling, multicollinearity remains a problem which may even be aggravated when using latent variable approaches. Further challenges of nonlinear latent analyses comprise the distribution of latent product terms, a problem especially relevant for approaches using maximum likelihood estimation methods based on multivariate normally distributed variables, and unbiased estimates of nonlinear effects under multicollinearity. The only methods that explicitly take the nonnormality of nonlinear latent models into account are latent moderated structural equations (LMS) and quasi-maximum likelihood (QML). In a small simulation study both methods yielded unbiased parameter estimates and correct estimates of standard errors for inferential statistic...","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"3 1","pages":"100-114"},"PeriodicalIF":3.1,"publicationDate":"2007-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1027/1614-2241.3.3.100","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292354","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：