Mike,
>
> So my primary concern is whether a minimum number of events is needed at
> EACH level of the parametric modulation term during first-level
(individual
> subject) analysis?
Given the assumption that model is correct, the estimator of the coefficient
of the modulator covariate will be unbiased regardless of how many
replications exist. However, the efficiency of the estimator can depend on
the structure of the modulator (such as how many replications exist), and
will ultimately be proportional to the norm of the covariate.
Eric
|