search

0todd0000 / spm1d

One-Dimensional Statistical Parametric Mapping in Python
GNU General Public License v3.0
61 stars 21 forks source link

GLM model assumption of normality #190

Closed eleonora-mont closed 3 years ago

eleonora-mont commented 3 years ago

Hi Todd,

I just run a normality test on my pressure data and this is what it looks like normality_test_39participants

The data is overall normally distributed although in some points as you can see, data crossed the threshold. Do you think it would be ok to use the GLM model on the present data? My concern is that it violetes the assumption of normality for the GLM model, although in few points.

Thanks,

Eleonora

0todd0000 commented 3 years ago

Hi Eleonora,

Are these 2D data? If yes, then only nonparametric analysis is possible in spm1d. It is fine to run these analyses (GLM or otherwise) because nonparametric procedures do not assume normality.

Todd

eleonora-mont commented 3 years ago

Hi Todd,

Thanks for your reply. Yes, this is 2D plantar pressure data. I thought that only parametric GLM procedure was available in spm1d, is this incorrect? That is why I run normality test to check on pressure data for it.

Thanks, Eleonora

0todd0000 commented 3 years ago

Yes, the GLM procedure is currently available only for parametric inference.

However, as discussed in #183, only nonparametric inference can be used for 2D analysis in spm1d. Thus there is no need to check normality for 2D analysis.

eleonora-mont commented 3 years ago

Yes true, I should have thought about it. This answers also to my initial question regarding the use of GLM to pressure data.

Thanks Todd.