Sorry for confusion!

I mean that in a fMRI experiment, we have used word as stimuli. Now, we do not care about what are the experimental conditions, we just want to reveal brain regions associated with one of the properties of the stimuli (word), such as the frequency of the word.

Because if we have several experimental conditions in a experiment, we will control the word frequency only between conditions, but not within condition. So, overall, words frequency in a experiment has a distribution. The other properties, such as Aoa (acquisition of age) is the same. If we want to examine which brain regions is associated with the particualr properties, such as the frequency, we can ignore the previous experimental condition, and do something like a multiple regression analysis. But for fMRI, I do not how to realize it. Can I replace the original stimuli, such as 0 0 0 1 1 1... with word frequency, such as 0 0 0 , 0.5, 0.7, 0.2...?

Thank you!