Dear Mozhdeh,
it is difficult to help if you don't provide any information about your kind of data, sample size and so on.
If you can't find any effects, there might be different reasons for that which are all linked to the idea of having enough statistical power to detect effects in your data:
1. There is no effect in your data.
2. Your sample size is too small to detect the effect with sufficient statistical power.
3. There is too much noise /variance in your data that minimize effect size.
Reason #3 is related to sample homogeneity and image quality, but also to potential confounds that might be added to your design (e.g. age, sex). And always check your data and quality parameters!
Best,
Christian
On Sat, 12 Jun 2021 10:01:46 +0430, Mozhdeh Haddadpour <[log in to unmask]> wrote:
>Thanks for responding. But now I am completely confused about how many
>times I should repeat "check sample homogeneity"???Because every time I
>exclude low quality data from my database and check the homogeneity of the
>remaining data again, I find that low quality data is still available.!!!
>Another question is about resampling size.Which value (32k or 164k) is
>suitable for resampling (also my data sets are public)?
>Thanks,
>Mozhdeh
>
|