Robust multilevel simultaneous component analysis

Abstract Multilevel simultaneous component analysis (MSCA) is a class of techniques for analyzing multivariate data with more than one level. However, as MSCA uses classical statistics such as the arithmetic mean, the resulting estimates are highly sensitive to outlying measurements. In this paper we propose a robust version of MSCA-P (the least restrictive MSCA variant) which can withstand the effects of outliers, while its associated outlier maps reveal the outliers visually. The technique is applied to data from an eating disorder study as well as to chemical data. We show by simulation that robust MSCA-P succeeds well in detecting outliers. The latter conclusion also holds when the data are generated according to the other MSCA variants.

[1]  P. Rousseeuw,et al.  A fast algorithm for the minimum covariance determinant estimator , 1999 .

[2]  Peter Filzmoser,et al.  A comparison of algorithms for the multivariate L1-median , 2010, Comput. Stat..

[3]  Frank Rijmen,et al.  Drive for thinness, affect regulation and physical activity in eating disorders: a daily life study. , 2007, Behaviour research and therapy.

[4]  Marieke E. Timmerman,et al.  Four simultaneous component models for the analysis of multivariate time series from more than one subject to model intraindividual and interindividual differences , 2003 .

[5]  Mia Hubert,et al.  LIBRA: a MATLAB library for robust analysis , 2005 .

[6]  Eva Ceulemans,et al.  Multilevel component analysis for studying intra-individual variability and inter-individual differences , 2009 .

[7]  Eva Ceulemans,et al.  A clusterwise simultaneous component method for capturing within-cluster differences in component variances and correlations. , 2013, The British journal of mathematical and statistical psychology.

[8]  Marieke E Timmerman,et al.  Multilevel component analysis. , 2006, The British journal of mathematical and statistical psychology.

[9]  Rasmus Bro,et al.  A phenomenological study of ripening of salted herring. Assessing homogeneity of data from different countries and laboratories , 2002 .

[10]  Age K. Smilde,et al.  Multilevel component analysis of time-resolved metabolic fingerprinting data , 2005 .

[11]  P. Rousseeuw.  Least Median of Squares Regression , 1984 .

[12]  Eva Ceulemans,et al.  Tolerance of Justice Violations: The Effects of Need on Emotional Reactions After Violating Equality in Social Dilemmas , 2011 .

[13]  Eva Ceulemans,et al.  Clusterwise simultaneous component analysis for analyzing structural differences in multivariate multiblock data. , 2012, Psychological methods.

[14]  Mia Hubert,et al.  ROBPCA: A New Approach to Robust Principal Component Analysis , 2005, Technometrics.

[15]  Age K Smilde,et al.  Bootstrap confidence intervals in multi-level simultaneous component analysis. , 2009, The British journal of mathematical and statistical psychology.

[16]  I. Mechelen,et al.  The co-occurrence of emotions in daily life: A multilevel approach , 2005 .

[17]  C. Croux,et al.  Generalizing univariate signed rank statistics for testing and estimating a multivariate location parameter , 1995 .

[18]  Peter Filzmoser,et al.  An Object-Oriented Framework for Robust Multivariate Analysis , 2009 .

[19]  E. Ceulemans,et al.  Universal Intracultural and Intercultural Dimensions of the Recalled Frequency of Emotional Experience , 2006 .

[20]  Peter J. Rousseeuw,et al.  Robust regression and outlier detection , 1987 .

[21]  Mia Hubert,et al.  Robustness and Outlier Detection in Chemometrics , 2006 .

[22]  Eva Ceulemans,et al.  How to perform multiblock component analysis in practice , 2011, Behavior Research Methods.

[23]  Rasmus Bro,et al.  Multi‐way models for sensory profiling data , 2008 .

[24]  Eva Ceulemans,et al.  The CHull procedure for selecting among multilevel component solutions , 2011 .

[25]  Onno E. de Noord,et al.  Multilevel component analysis and multilevel PLS of chemical process data , 2005 .