Hello, I investigate survival until the following year (0,1) and I wish to test if the variance in survival for two or more groups are significantly different from each other. I read that the Fligner-Killeen test is a non-parametric test which is very robust against departures from normality but is it correct (valuable technique for publication) to use it on binary data? In other words, can I use fligner.test(survival~categorical_predictor,data=mydata) when survival is binary (0,1)? Best regards Emeline --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus
Fligner-Killeen test on binary data
5 messages · Peter Dalgaard, emeline mourocq, Bert Gunter +1 more
That's not an R question but a stats question, but I wouldn't do it. For one thing: The variance of binary data is a function of the mean, so the research question is dubious in the first place. Secondly, the test is based on ranking and comparing absolute differences from the group median, which for binary data is generally 0 or 1, so all absolute differences will be 1.... Put differently, the results are more than likely to be complet rubbish. -pd
On 04 Apr 2016, at 11:48 , emeline mourocq <emeline.mourocq at uzh.ch> wrote: Hello, I investigate survival until the following year (0,1) and I wish to test if the variance in survival for two or more groups are significantly different from each other. I read that the Fligner-Killeen test is a non-parametric test which is very robust against departures from normality but is it correct (valuable technique for publication) to use it on binary data? In other words, can I use fligner.test(survival~categorical_predictor,data=mydata) when survival is binary (0,1)? Best regards Emeline --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus [[alternative HTML version deleted]]
______________________________________________ R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Peter Dalgaard, Professor, Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark Phone: (+45)38153501 Office: A 4.23 Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com
Thanks for all those information. I then have to find another way to test difference in variance between my two groups. -----Original Message----- From: peter dalgaard [mailto:pdalgd at gmail.com] Sent: Monday, 4 April, 2016 7:11 PM To: emeline mourocq <emeline.mourocq at uzh.ch> Cc: r-help at r-project.org Subject: Re: [R] Fligner-Killeen test on binary data That's not an R question but a stats question, but I wouldn't do it. For one thing: The variance of binary data is a function of the mean, so the research question is dubious in the first place. Secondly, the test is based on ranking and comparing absolute differences from the group median, which for binary data is generally 0 or 1, so all absolute differences will be 1.... Put differently, the results are more than likely to be complet rubbish. -pd
On 04 Apr 2016, at 11:48 , emeline mourocq <emeline.mourocq at uzh.ch> wrote: Hello, I investigate survival until the following year (0,1) and I wish to test if the variance in survival for two or more groups are significantly different from each other. I read that the Fligner-Killeen test is a non-parametric test which is very robust against departures from normality but is it correct (valuable technique for publication) to use it on binary data? In other words, can I use fligner.test(survival~categorical_predictor,data=mydata) when survival is binary (0,1)? Best regards Emeline --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus [[alternative HTML version deleted]]
______________________________________________ R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
-- Peter Dalgaard, Professor, Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark Phone: (+45)38153501 Office: A 4.23 Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus
Peter: Bravo!
On Mon, Apr 4, 2016 at 10:10 AM, peter dalgaard <pdalgd at gmail.com> wrote:
That's not an R question but a stats question, but I wouldn't do it. For one thing: The variance of binary data is a function of the mean, so the research question is dubious in the first place. Secondly, the test is based on ranking and comparing absolute differences from the group median, which for binary data is generally 0 or 1, so all absolute differences will be 1....
"Put differently, the results are more than likely to be complete rubbish." Again, probably not appropriate, but anyway: Fortune Nomination! Cheers, Bert
-pd
On 04 Apr 2016, at 11:48 , emeline mourocq <emeline.mourocq at uzh.ch> wrote: Hello, I investigate survival until the following year (0,1) and I wish to test if the variance in survival for two or more groups are significantly different from each other. I read that the Fligner-Killeen test is a non-parametric test which is very robust against departures from normality but is it correct (valuable technique for publication) to use it on binary data? In other words, can I use fligner.test(survival~categorical_predictor,data=mydata) when survival is binary (0,1)? Best regards Emeline --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus [[alternative HTML version deleted]]
______________________________________________ R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
-- Peter Dalgaard, Professor, Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark Phone: (+45)38153501 Office: A 4.23 Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com
______________________________________________ R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Hi emeline, I think there may be a minor language problem. If you mean the "variation" rather than the "variance" in survival, you may simply want a test of proportions. Jim
On Mon, Apr 4, 2016 at 7:48 PM, emeline mourocq <emeline.mourocq at uzh.ch> wrote:
Hello, I investigate survival until the following year (0,1) and I wish to test if the variance in survival for two or more groups are significantly different from each other. I read that the Fligner-Killeen test is a non-parametric test which is very robust against departures from normality but is it correct (valuable technique for publication) to use it on binary data? In other words, can I use fligner.test(survival~categorical_predictor,data=mydata) when survival is binary (0,1)? Best regards Emeline --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus [[alternative HTML version deleted]]
______________________________________________ R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.