Dear SPMers,
I was building a SPM model but kept on getting non-estimated beta in the design
matrix. In this model, I have three conditions (onset times for A, B, C) along
with the 1st order parametric modulation of RT from each condition. Condition B
consists of infrequent events; in some situations there is only one event in
this condition. The beta for the parametric modulation of RT in Condition B was
not estimated (i.e., in gray color) when Condition B had only one event, but it
was estimated when it had more than one event. Meanwhile, the beta for the
onset times for Condition B was estimated regardless of whether it had only one
or more events. Is it true that the parametric modulation needs at least two
data points in order to estimate the beta?
Thank you
Sien
|