Abstract
The detection of dependence structures within a set of random variables provides a valuable basis for a detailed subsequent investigation of their relationships. Nonparametric tests for independence require only basic assumptions on the marginal or joint distribution of the involved variables. In this paper, we review nonparametric tests of independence in bivariate as well as multivariate settings which are throughout ready-to-use, i.e., implemented in R packages. Highlighting their distinct empirical size and power properties in various small sample settings, our analysis supports an analyst in deciding for a most adequate test conditional on underlying distributional settings or data characteristics. Avoiding restrictive moment conditions, the copula based Cramér-von Mises distance of Genest and Rémillard (Test 13:335–370, 2004) is remarkably robust under the null hypothesis and powerful under diverse settings that are in line with the alternative hypothesis. Based on distinguished test outcomes in small samples, we detect nonlinear dependence structures between childhood malnutrition indices and possible determinants in an empirical application for India.
Similar content being viewed by others
Notes
Tests for serial dependence in time series are not explicitly considered here. An overview of corresponding approaches is given in Diks (2009).
More detailed descriptions of covered distributions and the theoretical background are given in Oja (2010). For the multivariate extension of Spearman’s rho we estimate the shape matrix by means of a rank based covariance matrix [see also the documentation of the corresponding R package SpatialNP (Sirkia et al. 2018)].
For instance, Shih and Emura (2016) study the properties of Spearman’s rho and Kendall’s tau under specific copula structures.
In this study, the considered nominal significance level is \(\alpha =0.05.\) Similar results obtain with respect to other conventional levels, for instance \(\alpha =0.1\).
Siqueira Santos et al. (2013) consider alternative choices for the distribution of \(x_1\) as, for instance, equidistant points or the uniform distribution. Additionally, they study further nonmonotonic and nonlinear dependence structures, i.e., alternative choices of the function f.
The asymptotic properties of Wilks’ Lambda have been shown under the multivariate Gaussian distribution. Thus, the comparison with the nonparametric tests is informative on the trade-off between efficient dependence detection within the Gaussian model, and robustness under more general distributional conditions.
References
Beran R, Bilodeau M, Lafaye de Micheaux P (2007) Nonparametric tests of independence between random vectors. J Multivar Anal 98(9):1805–1824
Blum JR, Kiefer J, Rosenblatt M (1961) Distribution free tests of independence based on the sample distribution function. Ann Math Stat 32:485–498
Bollerslev T (1990) Modelling the coherence in short-run nominal exchange rates: a multivariate generalized arch model. Rev Econ Stat 72:498–505
Deheuvels P (1981) An asymptotic decomposition for multivariate distribution-free tests of independence. J Multivar Anal 11(1):102–113
Diks C (2009) Nonparametric tests for independence. In: Meyers RA (ed) Encyclopedia of complexity and systems. Springer, New York, pp 6252–6271
Ding AA, Li Y (2014) Copula correlation: an equitable dependence measure and extension of pearson’s correlation. arXiv:1312.7214v3
Genest C, Quessy J-F, Rémillard B (2006) Local efficiency of a Cramér-von Mises test of independence. J Multivar Anal 97(1):274–294
Genest C, Quessy J-F, Remillard B (2007) Asymptotic local efficiency of Cramér-von Mises tests for multivariate independence. Ann Stat 35(1):166–191
Genest C, Rémillard B (2004) Tests of independence and randomness based on the empirical copula process. Test 13(2):335–370
Harrell FE (2015) Hmisc. R package version 3.15-0
Hoeffding W (1948) A non-parametric test of independence. Ann Math Stat 19:546–557
Hofert M, Kojadinovic I, Maechler M, Yan J (2015) Copula: multivariate dependence with copulas. R package version 0.999-13
Joe H (1997) Multivariate models and dependence concepts, vol 73. Monographs on statistics and applied probability. Chapman & Hall, London
Josse J, Holmes S (2014) Measures of dependence between random vectors and tests of independence. Literature review. arXiv:1307.7383v3
Kandala N, Lang S, Klasen S, Fahrmeir L (2001) Semiparametric analysis of the socio-demographic and spatial determinants of undernutrition in two african countries. Res Off Stat 1:81–100
Kendall MG (1938) A new measure of rank correlation. Biometrika 30(1/2):81–93
Klein N, Kneib T (2016) Simultaneous inference in structured additive conditional copula regression models: a unifying bayesian approach. Stat Comput 26(4):841–860
Kojadinovic I, Holmes M (2009) Tests of independence among continuous random vectors based on Cramér-von Mises functionals of the empirical copula process. J Multivar Anal 100(6):1137–1154
Manner H, Reznikova O (2012) A survey on time-varying copulas: specification, simulations, and application. Econ Rev 31(6):654–687
Matteson DS, Tsay RS (2013) Independent component analysis via distance covariance. Pre-print. arxiv.org/abs/1306.4911
Nakatani T (2010) Conditional Correlation GARCH Models. R package version (2):3
Oja H (2010) Multivariate nonparametric methods with R. In: Moller J (ed) An approach based on spatial signs and ranks, vol 199. Lecture Notes in statistics. Springer, New York
Pearson K (1920) Notes on the history of correlation. Biometrika 13(1):25–45
Puri ML, Sen PK (1971) Nonparametric methods in multivariate analysis. John Wiley & Sons Inc., New York
Rémillard B (2009) Discussion of: Brownian distance covariance. Ann Appl Stat 3(4):1295–1298
Rémillard B (2014) Tests of independence. In: Lovric M (ed) International encyclopedia of statistical science. Springer, Berlin, pp 1598–1601
Risk BB, James NA, Matteson DS (2015) steadyICA: ICA and tests of independence via multivariate distance covariance. R package version 1
Rizzo ML, Szekely GJ (2014) energy: E-statistics (energy statistics). R package version 1.6.2
Rosco J, Joe H (2013) Measures of tail asymmetry for bivariate copulas. Stat Pap 54(3):709–726
Sejdinovic D, Sriperumbudur B, Gretton A, Fukumizu K (2013) Equivalence of distance-based and RKHS-based statistics in hypothesis testing. Ann Stat 41(5):2263–2291
Shih J-H, Emura T (2016) Bivariate dependence measures and bivariate competing risks models under the generalized fgm copula. Stat Pap. https://doi.org/10.1007/s00362-016-0865-5
Siqueira Santos SD, Takahashi DY, Nakata A, Fujita A (2013) A comparative study of statistical methods used to identify dependencies between gene expression signals. Brief Bioinform 15(6):906–18
Sirkia S, Miettinen J, Nordhausen K, Oja H, Taskinen S (2018) SpatialNP: multivariate nonparametric methods based on spatial signs and ranks. R package version 1.1-3
Sklar M (1959) Fonctions de répartition à \(n\) dimensions et leurs marges. Publ Inst Stat Univ Paris 8:229–231
Spearman C (1904) The proof and measurement of association between two things. Am J Psychol 15:72–101
Székely GJ, Rizzo ML (2009) Brownian distance covariance. Ann Appl Stat 3(4):1236–1265
Székely GJ, Rizzo ML, Bakirov NK (2007) Measuring and testing dependence by correlation of distances. Ann Stat 35(6):2769–2794
Taskinen S, Oja H, Randles RH (2005) Multivariate nonparametric tests of independence. J Am Stat Assoc 100(471):916–925
Wilks S (1935) On the independence of \(k\) sets of normally distributed statistical variables. Econometrica 3(3):309–326
Acknowledgements
We thank two anonymous referees and the Associate Editor Christine H. Müller for helpful comments and suggestions. Financial support from the Academy of Finland (308628) is gratefully acknowledged.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Herwartz, H., Maxand, S. Nonparametric tests for independence: a review and comparative simulation study with an application to malnutrition data in India. Stat Papers 61, 2175–2201 (2020). https://doi.org/10.1007/s00362-018-1026-9
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00362-018-1026-9
Keywords
- Tests for independence
- Nonparametric methods
- Multivariate independence
- Spatial ranks
- Empirical copula
- Distance covariance