RESULTS: There were no reaction time effects that included smoking status nor was there an effect for accuracy on the 1-back task.
The factor run had no significant effect for accuracy and only showed a significant interaction with memory load for reaction time (RT) [ F 2,258) = 9.41, p < 0.001], indicating a decrease in RT for the 2-back condition only as the task progressed, although this effect did not differ across groups.
As is typically observed (Chepenik et al., 2007; van Steenbergen et al., 2010), mood did not have effects on interference or overall RT. 2 Moreover, no mood effects were found for accuracy, thus showing that the effect on conflict adaptation could not be attributed to a speed accuracy tradeoff.
No group effects were observed for accuracy.
Comparable effects were found for accuracy (Fig. 1 d, bottom left panel), with more errors occurring on trials with low mean than high mean [ F1,20 = 104, P < 0.001], or trials with high variance versus low variance [ F2,40 = 7.80, P < 0.001], and no interaction [ F2,40 = 2.70, P < 0.15].
On the other hand, these small numbers lead to small training sets, consequently with a detrimental effect for GS accuracy.
The mean absolute error for each joint angle and task are presented in Tables 1 and 2. The results from the repeat ANOVA did not yield a significant task × joint angle × matching arm interaction effect for kinesthetic accuracy (F2.27,111.27=0.69, P=0.52); however, the joint angle matching arm interaction effect was significant (F2.51,122.97= 8.28, P<0.05).
Mixed-model ANOVAs comparing the arcsine transformed accuracy rates and memory capacity estimates (K) as a function of load between Experiments 2a and 2b showed no effect for experiment, for accuracy rates, F 1, 27) = 1.91, p =.18, η =.07, and for K capacity estimates, F < 1, and importantly, no interaction (both Fs < 1).
