I would only model the null events if you think that the onset of the fixation cross causes a change in brain activity. In your case where the cross comes on after a blank ITI screen, I think I would model the cross. You still have blank time during the rest of the experiment by nature of the variable ITI.

Usually, the null-events are indistinguishable from the background and the subject is completely unaware of blank trials. This doesn't seem to be the case here where the "null events" are more like "catch" trials because the subject see the cross come on and is expecting a stimuli to follow.

Hope this helps.

Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Postdoctoral Research Fellow, GRECC, Bedford VA
Website: http://www.martinos.org/~mclaren
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
intended only for the use of the individual or entity named above. If the
reader of the e-mail is not the intended recipient or the employee or agent
responsible for delivering it to the intended recipient, you are hereby
notified that you are in possession of confidential and privileged
information. Any unauthorized use, disclosure, copying or the taking of any
action in reliance on the contents of this information is strictly
prohibited and may be unlawful. If you have received this e-mail
unintentionally, please immediately notify the sender via telephone at (773)
406-2464 or email.

On Mon, Mar 24, 2014 at 2:59 PM, H. Nebl <[log in to unmask]> wrote:

Dear experts,


as the headline implies, I wonder about whether or not to model null events in case of fast event-related designs. Sorry to bother you, as this is not directly related to SPM ;-) We have a fast event-related design with 4 conditions of interest A - D (brief visual stimuli) and another "null event" condition NE. During null events, a fixation cross shows up at first, similar to the conditions of interest, but no stimuli are presented after that. The ITI ranges from 1.2 to 3 s, with the mean roughly corresponding to our TR = 2 s. Overall we have about 600 trials, 50 of them are "null events".

Our main focus is on the differences between the 4 conditions. We get almost identical second-level results based on 1) first-level models without a NE regressor building contrasts A, B, C, D 2) first-level models including a NE regressor and building contrast vectors A - NE, B - NE, ...

However, I would also like to present the average activation pattern for our four conditions (A + B + C + D)/4. I'm aware that due to the fast event-related design the meaning of this pattern is somewhat limited, but I think it is nonetheless relevant enough to report (e.g. for future studies with similar designs). Depending on the models and contrasts, that is either (A + B + C + D)/4 or (A + B + C + D)/4 - NE, we get somewhat different (de)activations, that is smaller/larger clusters, although the general pattern remains the same.

I found a few papers on null events like Burock et al., 1998; Friston, Zarahn et al., 1999; Stark & Squire, 2001; Liu & Frank, 2004, but I didn't find an ultimative answer. Based on the rather low probability of our null events (~ 8 %) I tend not to model them. But I'm interested to hear some more opinions.


Thanks!

Helmut