How to draw decent conclusions with a unilateral region result?

#1

Hi NeuroStars community,

I am working on an experiment that using fMRI to investigate the involvement of the parahippocampal cortex (PHC) in spatial navigation.

From the results of GLM, I found a cluster in the left PHC which is significantly activated by spatial navigation condition.

I wonder how to draw a conclusion based on the unilateral region activity. Is it “the PHC is involved in spatial navigation”, or “the left PHC is involved in spatial navigation”? If the latter, should I interpret the non-significant result in the right side?

Any ideas would be appreciated.

Best,
Qunjun

#2

Hi,
You may want to test that the activity in left wd right PHC is significant. Unless it is, you cannot claim that there is a difference between left and right.
HTH,
Bertrand

1 Like
#3

Thank you Bertrand!

If I read your reply correctly, I can conclude that “the PHC is involved in spatial navigation” even the whole-brain analysis only finds the activity in the left side.

I supposed that the conclusion can only be drawn while the bilateral activations were observed.

Best,
Qunjun

#4

No, you can conclude “the left PHC is involved in spatial navigation”. I’m just saying that you can’t yet conclude on the difference between left and right, unless you test it explicitly.
Best,
Bertrand

#5

Thank you very much, I think I get your point.

But do you think is it also decent to say that “the PHC is involved in spatial navigation”? I found some previous studies drawing the generalized conclusions when they only observed the clusters at one side of the brain.

Best,
Qunjun

#6

No, there is evidence only for the left PHC. Neuroscientists are unfortunately not very rigorous in the way they report results…
Best,
Bertrand

1 Like
#7

You are right, the overstatements could be found here and there…

Thanks for being so patient with me.

Best,
Qunjun

#8

Just to echo Bertrand’s answer – you should explicitly test left vs right, one way to look at this is described here in section 7 https://onlinelibrary.wiley.com/doi/10.1111/ejn.14430: In the article we make the point that you should show the unthresholded stats map, and use lateralization indices (which test left vs right) to get a better understanding of how selective is the lateralized ‘significant’ response you observe.

cyril