Latent class models for clustering : a comparison with K-means

By Jay Magidson, Jeroen K. Vermunt. 2002

927 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/6add265688cde63766bed6b920c4546e7c11ab99

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

Latent Profile Analysis (2017, methodology): “Accordingly, we introduce LPA as a tool which is new to both family business and entrepreneurship research and holds considerable promise (Magidson & Vermunt, 2002, 2004).”
Disentangling Shape from Level Effects in Person-Centered Analyses: An Illustration Based on University Teachers’ Multidimensional Profiles of Effectiveness (2015, unknown): “…into mainstream psychological and educational research where they have superseded—perhaps due to their greater flexibility—cluster analytic (e.g., Magidson & Vermunt, 2002; Vermunt, 2011; but also see Steinley & Brusco, 2011) and taxometric methods (e.g., Lubke & Tueller, 2010; Waller &…”
What you see may not be what you get: Asking consumers what matters may not reflect what they choose (2010, result): “To take importance heterogeneity into account in the DCE results and test whether our findings of low importance for packaging attributes were simply due to aggregating over heterogeneous importances, we estimated a scale-extended latent class regression model (Magidson and Vermunt 2002) that simult”
Investigating population heterogeneity with factor mixture models. (2005, methodology): “LCA, LPA, and FMM, on the other hand, are model based (see Magidson & Vermunt, 2002, for a comparison of LCA methods to K-means clustering).”
Interrelations among university students’ approaches to learning, regulation of learning, and cognitive and attributional strategies: a person oriented approach (2011, methodology): “the more traditional clustering methods (Bacher et al. 2004; Magidson and Vermunt 2002), 520 High Educ (2011) 61:513–529”
Latent Class Models (2004, unknown): “While standardization often improves the situation, it does not solve the problem because equating the variance in the total sample is not the same as equating the within-group variances (Magidson and Vermunt, 2002).”
Modeling the behavioral determinants of travel behavior: An application of latent transition analysis (2014, background): “Additional benefits over cluster analysis are that statistical criteria can be used to judge the optimal number of classes and that variables of mixed-scale type can be accommodated (there is also no need to standardize variables) (Magidson and Vermunt, 2002).”
Subgroups of Alzheimer's disease based on cerebrospinal fluid molecular markers (2005, methodology): “Levels of total tau and A 1-42 were assayed by sandwich enzyme-linked immunosorbent assay using h Tau Ag and -Amyloid(1-42) kits (Innogenetics, Ghent, Belgium).”
A Multifoci Person-Centered Perspective on Workplace Affective Commitment: A Latent Profile/Factor Mixture Analysis (2011, background): “For a detailed comparison of cluster analyses and LPA or related models, the interested reader is referred to Magidson and Vermunt (2002).”
Latent Class Analysis (2021, methodology): “In comparison to LC, K-means has been shown to perform poorly in cluster recovery because the K-means assumption that residual variances are equal within each cluster fails to hold true in practice (Magidson & Vermunt, 2002a, 2002b; Vermunt, 2011).”
Relations between teacher students’ approaches to learning, cognitive and attributional strategies, well-being, and study success (2012, result): “Note, that although recent simulation and method comparison studies provide evidence supporting the use of LCCA instead of the more traditional clustering methods (Bacher et al. 2004; Magidson and Vermunt 2002), justified criticism directed at the Practical Value of some of the above advantages and ”
Measuring consumers attitudes towards health and taste and their association with food-related life-styles and preferences (2019, methodology): “The appropriateness of the selected model was further evaluated through the size of the bivariate residual, which did not exceed the critical value of 1 (Vermount & Magidson, 2002).”