The robustness of LISREL modeling revisted.

By Anne Boomsma, J. J. Hoogland. 2001

548 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ae7a437f4b3c766aa3ca98442faf6bc19239eb6e

How researchers see this paper

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

Citations in the corpus

A parent-report measure of children's anxiety: psychometric properties and comparison with child-report in a clinic and normal sample. (2004, methodology): “Because of the skewness and kurtosis in the data, we choose the parameter estimation that was most robust to non-normality: the robust maximum likelihood parameter using an asymptotic covariance matrix (Boomsma & Hoogland, 2001).”
Applied Psychometrics: Sample Size and Sample Power Considerations in Factor Analysis (EFA, CFA) and SEM in General (2018, methodology): “…1999; Hoyle & Kenny, 1999; Marsh & Hau, 1999), but generally N = 100 - 150 is set as a minimum sample size for SEM research (Anderson & Gerbing, 1988; Ding, Velicer, & Harlow, 1995) while others set this minimum to N = 200 (Hoogland & Boomsma, 1998; Boomsma & Hoogland, 2001) as per Loehlin (2004).”
Antecedents and consequences of employee absenteeism: A longitudinal perspective on the role of job satisfaction and burnout (2010, methodology): “RML is appropriate for variables with nonnormal distributions (Boomsma & Hoogland, 2001).”
Students’ perceptions of teaching quality in higher education: the perspective of currently enrolled students (2007, methodology): “To take account of this violation of assumptions, the confirmatory factor analysis used robust maximum likelihood estimation, a method less sensitive to violations of the normality assumption with large models than other estimation methods (Boomsma & Hoogland, 2001).”
Body Image and Self-Esteem Among Adolescent Girls: Testing the Influence of Sociocultural Factors (2005, background): “However, we encountered no problems of nonconvergence or improper solutions—common difficulties with small sample sizes in structural equation modeling (Boomsma & Hoogland, 2001)—and the analysis had sufficient power to detect the effects of interest.”
Teachers'motivation to participate in training and to implement innovations (2014, methodology): “Our analyses comply with recommendations that when sample size is small, it may be appropriate to increase indicators per factor ratio, so in our models we constructed two latent variables with six indicators each (Boomsma & Hoogland, 2001; Marsh & Hau, 1999; Marsh, Hau, Balla, & Grayson, 1998).”
PLS-SEM: Looking Back and Moving Forward (2014, methodology): “In contrast, CB-SEM in similar situations has several notable weaknesses, including a lack of robustness (Boomsma and Hoogland, 2001). The obvious misapplications of PLS-SEM in many past research articles resulted in the publications of guidelines for researchers, reviewers and editors on how to con”
A Longitudinal Model of Continued IS Use: An Integrative View of Four Mechanisms Underlying Postadoption Phenomena (2005, background): “Thus, common problems resulting from small sample size, e.g., nonconvergence and negative variance estimates, may make it difficult to test these intermediate models (Boomsma and Hoogland 2001).”
Developing Attitudes Toward an Entrepreneurial Career Through Mentoring (2015, methodology): “We performed a CFA on all the variables used in this study and the model fit indices show good validity (w2 ¼ 254.9111, df ¼ 164, root mean square error of approximation [RMSEA] ¼ 0.04672, comparative fit index [CFI] ¼ 0.9839; Boomsma & Hoogland, 2001).”
When to use and how to report the results of PLS-SEM (2019, background): “, 2000), although it may require much larger sample sizes (Boomsma and Hoogland, 2001).”
The effects of service quality, image, and customer satisfaction on customer complaints and loyalty in high-speed rail service in Turkey: a proposal of the structural equation model (2017, methodology): “As simulation studies demonstrate, robust ML estimators based on the Satorra–Bentler scaled χ2 statistic have relatively good statistical properties compared to LS estimators (Hoogland 1999; Boomsma and Hoogland 2001).”
Testing for multigroup equivalence of a measuring instrument: a walk through the process. (2008, methodology): “For more detailed descriptions of tests for multigroup equivalence, readers are referred to Horn & McArdle (1992), Little (1997), and Widaman and Reise (1997); to Byrne (1998, 2001, 2006) for annotated explanation and illustration of diverse models based on the LISREL, AMOS, and EQS programs, respec”
Assessing Factorial Invariance in Ordered-Categorical Measures (2004, methodology): “While WLS estimation was used in the examples, this method is known to require very large samples for adequate performance (Boomsma & Hoogland, 2001; Muthén & Kaplan, 1985).”