108 / 218
106
Chapter 5
task switching, we observed slower and less accurate responding on switch than repeat trials.
The switch cost in terms of accuracy (i.e. less accurate responding on switch vs. repeat trials)
was larger than the switch cost in terms of response times. This resulted in a switch-related
increase in the speed-over-accuracy score. Moreover, we observed a reward-related shift in
the SAT during task switching: A reward-related increase in speed-over-accuracy on the more
demanding switch trials, but a reward-related decrease in speed-over-accuracy on the easier
repeat trials.
Effects of age on rewarded task switching
The aim of the current study was to look at age-related changes in the motivational
enhancement of cognitive control. Before assessing this primary effect of interest, we first
looked at age-related changes in the main effects of Reward and Task switching. In terms of
the SAT we observed an age-related decrease in speed over accuracy in terms of the effect of
Reward (Reward x Age: r (ρ) = -0. 191, p = 0.038) and of Task switching (Task switching x
age: r (ρ) = -0.225, p = 0.014). This age-related decrease in speed over accuracy as a function
of reward was due to a larger age-related increase in response times than accuracy on high
compared with low reward trials. In terms of task switching, younger participants exhibit a
relatively larger switch cost in terms of accuracy than in terms of speed. With age, this switch
cost in terms of speed increased, whereas the switch cost in terms of accuracy decreased.
Next, we assessed age-related changes in the main effects of Reward and Task switching in
terms of response times and accuracy and observed an age-related decrease in the difference
between high versus low reward in terms of accuracy (Age x Reward (high – low): r (ρ) =
-0.238, p = 0.009), and response times (Age x Reward (high – low): r (ρ) = 0.191, p = 0.038).
In addition, there was an age-related increase in slowing and accuracy on switch versus repeat
trials (RTs: Age x Task switching (switch – repeat): r (ρ) = 0.411, p < 0.001; accuracy: Age x
Task switching (switch – repeat): r (ρ) = 0.302, p < 0.001).
Thus aging was accompanied by a reduced reward benefit in terms of accuracy and an
increased reward impairment in terms of response times. A larger age-related increase in
response times than accuracy on high compared with low reward trials resulted in the age-
related decrease in speed over accuracy. In addition, there was an age-related increase in
task switching cost in terms of response times and an age-related switch benefit in terms of
accuracy, but this switch cost in terms of response times changed more with age, resulting in
an age-related decrease in speed over accuracy.
Next, we proceeded to our primary question of interest and assessed the degree to which a
promised Reward affected Task switching in terms of SAT and found that this effect changed
with age (Age x Reward x Task switching: r (ρ) = - 0.368, p < 0.001;
figure 5.2a
). Breaking
down this three-way Age x Reward x Task switching interaction for repeat and switch trials
separately revealed a positive correlation between Age and the Reward effect on repeat trials
(Age x Reward: r (ρ) = 0.311, p = 0.001), but a negative correlation betweenAge and the Reward