Effect of the Number of Variables on Measures of Fit in Structural Equation Modeling

By David A. Kenny, Betsy McCoach. 2003

1,201 citations recorded by the source corpus.

Subfields: Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/1afe3760ce8b1231571a20dc36f7f2e8b847510d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Management Science and Operations Research

Citations in the corpus

The paths to social licence to operate: An integrative model explaining community acceptance of mining $ (2014, methodology): “A satisfactory fit is indicated by a nonsignificant chi-square test, CFIZ .95, NFIZ .95, and RMSEAr .06 (Hu and Bentler, 1999; Kenny and McCoach, 2003).”
The ubiquity of common method variance: The case of the Big Five (2011, methodology): “There is much evidence that combining items into parcels results in models that exhibit better goodness-of-fit statistics than models of the same data using items as indicators (e.g., Kenny & McCoach, 2003; Lim & Ployhart, 2006; Thompson & Melancon, 1996).”
Assessing mindfulness in children and adolescents: development and validation of the Child and Adolescent Mindfulness Measure (CAMM). (2011, methodology): “This pattern is consistent with the findings of Kenny and McCoach (2003), who reported that, even in correctly specified models with strong absolute fit index values, relative fit indices, particularly the NNFI, tend to decrease when there are many items (more than five or six).”
Effects of general and broad cognitive abilities on mathematics achievement. (2008, background): “The TLI and CFI are also reported; however, these indexes tend to worsen when models contain many different variables, as in the current study (Kenny & McCoach, 2003).”
An Empirical Evaluation of the Use of Fixed Cutoff Points in RMSEA Test Statistic in Structural Equation Models (2008, methodology): “There are a number of well-designed Monte Carlo simulation studies examining the performance of SEM fit indices, including the RMSEA (Hu and Bentler 1998; Fan, Thompson, and Wang 1999; Kenny and McCoach 2003; Nasser and Wisenbaker 2003).”
Refining the theory of basic individual values. (2012, background): “As Kenny and McCoach (2003) noted, even in correctly specified models, CFI tends to worsen as the number of variables in a model grows large.”
Resisting Everything except Temptation: Evidence and an Explanation for Domain–specific Impulsivity (2012, methodology): “90 indicate a good fit; however, Kenny and McCoach (2003) note that ‘the CFI tend[s] to demonstrate worse fit as the number of variables in the model increases.”
Revisiting the Model Size Effect in Structural Equation Modeling (2018, background): “For example, Kenny and McCoach (2003) found that when the sample size is small (e.g., less than 200), the empirical distribution of the likelihood ratio test statistics exhibits greater upward bias2 as the number of observed variables (p) increases, even when the models are correctly specified.”
Measuring The Change in Faculty Perceptions Over Time: An Examination of Their Worklife and Satisfaction (2005, methodology): “In these three preliminary steps, the fit of the proposed structural models to these data can be assessed with several commonly used indices ( Kenny and McCoach, 2003 ): the root-mean-square error of approximation (RMSEA), the chi-square test (v 2 ), the standardized root-mean-square”
Assessing Fit in Ordinal Factor Analysis Models: SRMR vs. RMSEA (2020, methodology): “…has examined the performance of point estimates, confidence intervals and tests of close fit using the RMSEA (Curran, Bollen, Chen, Paxton, & Kirby, 2003; Fan, Thompson, & Wang, 1999; Hu & Bentler, 1998; Kenny et al., 2015; Kenny & McCoach, 2003; Nasser & Wisenbaker, 2003; Nevitt & Hancock, 2000).”
Examining Chi-Square Test Statistics Under Conditions of Large Model Size and Ordinal Data (2018, background): “…have noted a model size effect with inflated chi-square statistics and increasing Type I error as model size increased (Anderson & Gerbing, 1984; Beauducel & Herzberg, 2006; Herzog et al., 2007; Kenny & McCoach, 2003; Marsh, Hau, Balla, & Grayson, 1998; Shi, Lee, & Terry, 2018; Shi et al., 2015).”
Age Diversity, Age Discrimination Climate and Performance Consequences - a Cross Organizational Study (2011, background): “32, 264–290 (2011) DOI: 10.1002/job size below 200 (Chen, Curra, Bollen, Kirby, & Paxton, 2008; Sharma et al., 2005; Hu & Bentler, 1999) and also depends on the number of variables in the model (Kenny & McCoach, 2003).”
Partner Reactions to Work-to-Family Conflict: Cognitive Appraisal and Indirect Crossover in Couples (2011, methodology): “As per discussion of goodness-of-fit rules of thumb (see Kenny & McCoach, 2003; Marsh, 2004) we used the cut-offs of .90 for CFI and .08 for RMSEA given our small sample size and number of variables.”