Thanks Donald and Colin. Yes, "null event" is somewhat misleading in that context. The term has been used in similar studies though, probably due to the stimuli involved. Some of them are "targets" and others are "no targets". Thus it would be confusing to call the "null event" a "no target" event, and "no stimulus" might also be misleading, maybe "fixation cross only" ;-)
Due to the design the "null events" are unlikely to act as catch trials insofar as they do not (seem to) provoke a motor repsonse on behavioral level. However, it is likely that the appearance of the fixation cross leads to some reorienting/other attentional mechanisms. This would actually be worth while to control for with contrasts like A - NE. On the other hand, due to their low probability the "null events" might indeed result in additional activations in the sense of violation of expectation.
But, as I said, the focus is on the differences between A, B, C, D, and here modelling/taking into account null events has no influence at all. Seems Colin's suggestion is the most convincing approach then. Probably it would have been best to either increase the probability of the "null events" to avoid anything related (then it might serve as a proper control condition), or alternatively, not to present any fixation cross at all (then there's nothing to model).
|