Drexel dragonThe Math ForumDonate to the Math Forum



Search All of the Math Forum:

Views expressed in these public forums are not endorsed by Drexel University or The Math Forum.


Math Forum » Discussions » sci.math.* » sci.stat.math.independent

Topic: one-way repeated measures ANOVA with more levels than subjects
Replies: 8   Last Post: Nov 6, 2013 3:53 AM

Advanced Search

Back to Topic List Back to Topic List Jump to Tree View Jump to Tree View   Messages: [ Previous | Next ]
Tim

Posts: 5
Registered: 10/23/13
Re: one-way repeated measures ANOVA with more levels than subjects
Posted: Oct 23, 2013 4:56 PM
  Click to see the message monospaced in plain text Plain Text   Click to reply to this topic Reply

Hi Bruce,

Thanks a lot for your answer. As you said, SPSS did not produce output for the Multivariate Tests.

Sorry for originally not being more explicit about the function we are using, when I posted I was generally wondering whether there was a principled reason why the repeated measures ANOVA should not work properly in case of more levels than subjects.

We are in fact using the statfun_depsamplesF.m function of the fieldtrip package in matlab (we are looking at amplitudes of electroencephalography data). I will dive into the details tomorrow and try to track down whether it runs a MANOVA internally. Thanks for the hint, I will keep you updated.

Tim



On Wednesday, October 23, 2013 7:32:03 PM UTC+2, Bruce Weaver wrote:
> On 23/10/2013 12:48 PM, Tim wrote:
>

> > Hi everyone,
>
> >
>
> > I have a quick question about a one-way repeated measures ANOVA in cases where the number of levels in a factor is larger than the number of subjects measured. Let me illustrate my question with a quick example: Say we presented 20 different colors in an experiment and recorded reaction times for every condition in 15 subjects (a total of 20*15=300 measurements). To my understanding, as we have 15 within-subject measurements for every cell, this should work perfectly fine.
>
> >
>
> > Unfortunately, two of the statistics functions we tried using throw an error stating that the number of conditions is too large given the number of subjects (one seems to directly compare the two numbers, the other one computes negative df_error). However, the same kind of setting works just fine in SPSS (df1 = 19; df2 = 266).
>
> >
>
> > Is there any principled reason why above setting should be problematic in a repeated-measures ANOVA? My search for an explanation has so far not brought up any reason why the number of levels in a factor should not exceed the number of subjects if all observations have been measured (n_obs = n_sub * n_cond).
>
> >
>
> >
>
> > Thanks a lot in advance, looking forward to your answers
>
> > Tim
>
> >
>
>
>
> If you ran this with SPSS, you will have seen that it does not produce
>
> any results in the "Multivariate Tests" table, and a footnote gives this
>
> explanation:
>
>
>
> "Cannot produce multivariate test statistics because of insufficient
>
> residual degrees of freedom."
>
>
>
> This makes me wonder if the two "statistics functions" you refer to are
>
> attempting to carry out MANOVA rather than repeated measures ANOVA.
>
> Perhaps you could provide more info about those "functions".
>
>
>
> HTH.
>
>
>
> --
>
> Bruce Weaver
>
> bweaver@lakeheadu.ca
>
> http://sites.google.com/a/lakeheadu.ca/bweaver/Home
>
> "When all else fails, RTFM."





Point your RSS reader here for a feed of the latest messages in this topic.

[Privacy Policy] [Terms of Use]

© Drexel University 1994-2014. All Rights Reserved.
The Math Forum is a research and educational enterprise of the Drexel University School of Education.