Regression slopes

[Ieva Andrulyte]

Hi Frank,

I have been using DSI Studio for a while and have found some interesting results with continuous analysis, including both positive and negative correlations. I wanted to better understand the associations of each tract, so I clustered the significant tracts and extracted QA statistics from each tract for each participant to do separate regression analyses and plot them. However, when I plot each significant part of the different tracts, there is no significant correlation when looking at the individual subjects. Could you help me better understand why this is the case, considering those tracts were significant (after FDR correction) when running connectometry?

Many thanks,
Ieva

[Frank Yeh]

There could be some possible causes/solutions to this problem:

1. low T-threshold: using a lower T threshold (e.g. 2) may give a weak correlation even if the FDR is very small. Increasing the T-threshold should help.

2. small sample size: if your number of subjects are less than 20, then it is likely that there is a strong bias. There is not much we can do if the subject number is not enough.

3. I would suggest updating DSI Studio to the most recent version (even if it is just a few months ago).

--
You received this message because you are subscribed to the Google Groups "DSI Studio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dsi-studio+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/1d9bfe86-91e9-4cf3-bf60-01ee39f6ad10n%40googlegroups.com.