Between and Within Subject Effect

[Xuefei Pan]

I want to test whether the Drift rates (V) was significantly different from each other for two conditions (‘rhythm’, ‘random’) which was  within-subject designed.

With the function hddm.HDDM ,the possibility of V_rhythm > V_random was bigger than 95% , meaning a significant difference of V between the two conditions.

However, with the function hddm.HDDMRegressor , a method testing the within-subject effect in the HDDM tutorial, the posterior of the beta estimated in the regression model was not significant different with 0. 

In many statistic case, with the same data, the within-subject effect is bigger than the between-subject effect. I am wondering whether there was something wrong with my analysis or it is a possible situation that may occurred in the modeling.

[valer...@gmail.com]

Hi,

did you manage to get your head around this?

I am trying to achieve something similar.

Best wishes,

Valerio

[hcp...@gmail.com]

There was once a thread about this: https://groups.google.com/g/hddm-users/c/eJ3vsuR4rRE/m/ZEUZQP4kBAAJ