[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: RE: Non-parametric tests for survey data? (e.g., Kruskal-Wallace)

From   "Michael I. Lichter" <>
Subject   Re: st: RE: Non-parametric tests for survey data? (e.g., Kruskal-Wallace)
Date   Wed, 11 Feb 2009 13:13:19 -0500

Thank you Nick and Stas for clarifying the reasons behind the lack of nonparametric procedures for complex survey data.

Thank you Roger for suggesting Somers' D. What I glean from the documentation suggests that the X variable in -somersd- can be dichotomous or it can be at least ordinal, but it cannot be nominal with k > 2 categories. For my (4-category nominal) situation, I would need to do multiple tests, possibly with a Bonferroni adjustment. That sounds doable.

Stas, I'd be interested on your take on the application of Somer's D to complex survey data. -somersd- allows jackknifed errors, which should relax the assumption of independence, and the statistics is supposed to be robust to distributional differences, and has a well-defined population expectation.

Roger, FWIW, the help page for somersd says that your dissertation was completed in 1887, and " {help lincom]" should, I guess, be "{help lincom}".



Newson, Roger B wrote:
One possibility might be to use the -somersd- package, downloadable from
SSC using the -ssc- command in Stata. The -somersd- package generates
confidence intervals for a wide range of rank statistics (particularly
Somers' D and Kendall's tau-a), and can be used with
sampling-probability weights (pweights) and/or the -svy:- prefix. It
comes with 3 .pdf manuals, which you can get when you download the

I hope this helps.

Best wishes


Roger B Newson BSc MSc DPhil
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton Campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: Web page:
Departmental Web page:

Opinions expressed are those of the author, not of the institution.

-----Original Message-----
[] On Behalf Of Michael I.
Sent: 10 February 2009 21:25
Subject: st: Non-parametric tests for survey data? (e.g.,

I don't see any procedures for doing non-parametric tests (aside from chi-square in svy: tab) with complex survey data (stratified, unequal probabilities of selection). I am particularly looking for tests of difference in ordinal dependent variables across k groups (k > 2). Kruskall-Wallace is the most obvious test, but only available for non-survey data. I assume that these procedures are not available because (a) it's not clear what to do with weights in nonparametric analyses anyway (which I infer partly from the fact that none of Stata's nonparametric procedures

take weights), (b) because there's no theory about whether/how they should work, and/or (c) because nobody has gotten around to it yet.

I'm looking for suggestions.

One possibility that comes to mind is to generate ranks using -egen- and

analyze using -svy: mean- or -svy: reg- (I'd use one-way ANOVA if somebody could explain how to do it with -svy- commands). I could also do -svy: intreg- for the variables that represent ranges underlying continuous variables (since most of my ordinal variables do represent well-defined but unequal-sized ranges of underlying continuous variables, e.g., 1 = "> 1", 2 = "2-4" 3 = "5 or more"), but that would require -intreg- to be robust to floor effects, and I doubt that it is (since the method assumes an underlying Normal distribution). (I guess -mlogit-, -ologit- and -gologit2- are also possibilities.)


Michael I. Lichter, Ph.D.
Research Assistant Professor & NRSA Fellow
UB Department of Family Medicine / Primary Care Research Institute
UB Clinical Center, 462 Grider Street, Buffalo, NY 14215
Office: CC 125 / Phone: 716-898-4751 / E-Mail:

*   For searches and help try:

© Copyright 1996–2021 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index