Abstract
In a one-way design with unknown and unequal error variances, we are interested in testing the null hypothesis \(H_0:\mu _1=\mu _2=\cdots =\mu _k\) against the ordered alternative \(H_1:\mu _1\le \mu _2\le \cdots \le \mu _k\) (with at least one strict inequality). We propose the likelihood ratio test (LRT) as well as two Min-T tests (Min-T and AMin-T) and study their asymptotic behaviour. For better finite sample properties, a parametric bootstrap procedure is used. Asymptotic accuracy of the parametric bootstrap is established. Simulation results show that all the tests control the nominal level in case of small, moderate, and highly unbalanced sample sizes. This even holds for combinations with large variances. However, the asymptotic likelihood ratio test (ALRT) achieves the nominal size only for large samples and \(k=3\). Power comparisons between LRT and Min-T test show that the LRT has more power than Min-T. However, the Min-T test is easier to use as it does not require evaluation of MLEs. Similar observations are made while comparing ALRT and AMin-T test. The robustness of these tests under departure from normality is also investigated by considering five non-normal distributions (t-distribution, Laplace, exponential, Weibull, and lognormal). When high variances are combined with small or highly unbalanced sample sizes, the tests are observed to be robust only for symmetric distributions. If variances of groups are equal and sample sizes of each group are at least 10, the proposed tests perform satisfactorily for symmetric and skewed distributions. R packages are developed for all tests and a practical example is provided to illustrate their application.
Similar content being viewed by others
References
Alvandi SSM, Jafari AA (2013) A parametric bootstrap approach for one-way ANCOVA with unequal variances. Commun Stat-Theory Methods 42(14):2473–2498
Bartholomew DJ (1959a) A test of homogeneity for ordered alternatives. Biometrika 46(1/2):36–48
Bartholomew DJ (1959b) A test of homogeneity for ordered alternatives II. Biometrika 46(3/4):328–335
Bickel PJ, Freedman DA (1981) Some asymptotic theory for the bootstrap. Ann Stat 9(6):1196–1217
Bishop TA, Dudewicz EJ (1978) Exact analysis of variance with unequal variances: test procedures and tables. Technometrics 20(4):419–430
Bishop TA, Dudewicz EJ (1981) Heteroscedastic ANOVA. Sankhya 43B:40–57
Bretz F (2006) An extension of the Williams trend test to general unbalanced linear models. Comput Stat Data Anal 50(7):1735–1748
Cabilio P, Peng J (2008) Multiple rank-based testing for ordered alternatives with incomplete data. Stat Probab Lett 78(16):2609–2613
Chen SY (2000) One-sided range test for testing against an ordered alternative under heteroscedasticity. Commun Stat-Simul Comput 29(4):1255–1272
Chen SY (2001) One-stage and two-stage statistical inference under heteroscedasticity. Commun Stat-Simul Comput 30(4):991–1009
Chang C, Pal N (2008) Testing on the common mean of several normal distributions. Comput Stat Data Anal 53(2):321–333
Cohen A, Kemperman JHB, Sackrowitz HB (2000) Properties of likelihood inference for order restricted models. J Multivar Anal 72(1):50–77
Donaldson T (1968) Robustness of the F-test to errors of both kinds and the correlation between the numerator and denominator of the F-ratio. J Am Stat Assoc 63(322):660–676
Hasler M (2014) Multiple contrast tests for multiple endpoints in the presence of heteroscedasticity. Int J Biostat 10(1):17–28
Herberich E, Sikorski J, Hothorn T (2010) A robust procedure for comparing multiple means under heteroscedasticity in unbalanced designs. PLoS ONE 5(3):e9788
Hayter AJ (1990) A one-sided studentized range test for testing against a simple ordered alternative. J Am Stat Assoc 85(411):778–785
Hayter AJ, Liu W (1996) Exact calculations for the one-sided studentized range test for testing against a simple ordered alternative. Comput Stat Data Anal 22(1):17–25
Jonckheere AR (1954) A test of significance for the relation between \(m\) rankings and \(k\) ranked categories. Br J Stat Psychol 7(2):93–100
Kassahun T, Wencheko E, Bathke AC (2021) A rank-based test for monotone trends in k-sample multivariate problems. Commun Stat-Theory Methods 1–22
Konietschke F, Bösiger S, Brunner E, Hothorn LA (2013) Are multiple contrast tests superior to the ANOVA? Int J Biostat 9(1):63–73
Konietschke F, Bathke AC, Harrar SW, Pauly M (2015) Parametric and non parametric bootstrap methods for general MANOVA. J Multivar Anal 140:291–301
Malekzadeh A, Kharrati-Kopaei M (2018) Inferences on the common mean of several normal populations under heteroscedasticity. Comput Stat 33(3):1367–1384
Marcus R (1976) The powers of some tests of the equality of normal means against an ordered alternative. Biometrika 5(1):37–42
Noguchi K, Konietschke F, Marmolejo-Ramos F, Pauly M (2021) Permutation tests are robust and powerful at 0.5% and 5% significance levels. Behav Res Methods 53(6):2712–2724
Pal N, Linb J, Changc C, Kumar S (2007) A revisit to the common mean problem: comparing the maximum likelihood estimator with the Graybill-Deal estimator. Comput Stat Data Anal 51(12):5673–5681
Pallman P, Hothron LA (2016) Analysis of means: a generalized approach using R. J Appl Stat 43(8):1541–1560
Pauly M, Brunner E, Konietschke F (2015) Asymptotic permutation tests in general factorial designs. J R Stat Soc Ser B 77(2):461–473
Pauly M, Asendorf T, Konietschke F (2016) Permutation-based inference for the AUC: a unified approach for continuous and discontinuous data. Biom J 58(6):1319–1337
Ramseyer GC, Tcheng TK (1973) The robustness of the studentized range statistic to violations of the normality and homogeneity of variance assumptions. Am Educ Res J 10(3):235–240
Shan G, Young D, Kang L (2014) A new powerful nonparametric rank test for ordered alternative problem. PLoS ONE 9(11):e112924
Shi NZ, Jiang H (1998) Maximum likelihood estimation of isotonic normal means with unknown variances. J Multivar Anal 64(2):183–195
Shirley E (1997) A non-parametric equivalent of Williams’ test for contrasting increasing dose levels of a treatment. Biometrics 33:386–389
Srivastava ABL (1958) Effect of non-normality on the power function of t-test. Biometrika 45(3/4):421–430
Williams DA (1971) A test for differences between treatment means when several dose levels are compared with a zero dose control. Biometrics 27(1):103–117
Williams DA (1972) The comparison of several dose levels with a zero dose control. Biometrics 28:519–531
Acknowledgements
The authors thank two referees and the chief editor for their comments and constructive suggestions which has helped in improving the manuscript substantially.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Mondal, A., Pauly, M. & Kumar, S. Testing for ordered alternatives in heteroscedastic ANOVA under normality. Stat Papers 64, 1913–1941 (2023). https://doi.org/10.1007/s00362-022-01366-0
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00362-022-01366-0