Limits...
Post-error slowing in sequential action: an aging study.

Ruitenberg MF, Abrahamse EL, De Kleine E, Verwey WB - Front Psychol (2014)

Bottom Line: Importantly, for young adults and middle-aged participants the observed slowing was also accompanied by increased accuracy after an error.We suggest that slowing in these age groups involves both functional and non-functional components, while slowing in elderly participants is non-functional.Moreover, using action sequences (instead of single key-presses) may allow to better track the effects on performance of making an error.

View Article: PubMed Central - PubMed

Affiliation: Department of Cognitive Psychology and Ergonomics, University of Twente Enschede, Netherlands.

ABSTRACT
PREVIOUS STUDIES DEMONSTRATED SIGNIFICANT DIFFERENCES IN THE LEARNING AND PERFORMANCE OF DISCRETE MOVEMENT SEQUENCES ACROSS THE LIFESPAN: Young adults (18-28 years) showed more indications for the development of (implicit) motor chunks and explicit sequence knowledge than middle-aged (55-62 years; Verwey et al., 2011) and elderly participants (75-88 years; Verwey, 2010). Still, even in the absence of indications for motor chunks, the middle-aged and elderly participants showed some performance improvement too. This was attributed to a sequence learning mechanism in which individual reactions are primed by implicit sequential knowledge. The present work further examined sequential movement skill across these age groups. We explored the consequences of making an error on the execution of a subsequent sequence, and investigated whether this is modulated by aging. To that end, we re-analyzed the data from our previous studies. Results demonstrate that sequencing performance is slowed after an error has been made in the previous sequence. Importantly, for young adults and middle-aged participants the observed slowing was also accompanied by increased accuracy after an error. We suggest that slowing in these age groups involves both functional and non-functional components, while slowing in elderly participants is non-functional. Moreover, using action sequences (instead of single key-presses) may allow to better track the effects on performance of making an error.

No MeSH data available.


Related in: MedlinePlus

Mean RTs per key press within post-error (open circles) and post-correct (black squares) 6-key sequences as a function of age group. Error bars represent standard errors.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3927119&req=5

Figure 2: Mean RTs per key press within post-error (open circles) and post-correct (black squares) 6-key sequences as a function of age group. Error bars represent standard errors.

Mentions: For the 6-key sequence, the data of 23 young adults, 20 middle-aged and 23 elderly participants were included. Results of the ANOVA showed that post-error sequences were generally performed slower than post-correct sequences (714 vs. 612 ms), F(1, 60) = 28.28, p < 0.001, η2p = 0.32. So, the average post-error sequence slowing amounted to 102 ms. An additional Trial type x Key interaction suggested that the amount of post-error slowing differed between the key presses within the sequence, F(5, 300) = 4.00, p < 0.05, η2p = 0.06. Further analyses showed that the interaction was no longer significant after removing the first key press of the sequence from the analysis (p = 0.37), while the main effect of Trial type still remained significant, F(1, 60) = 23.52, p < 0.001, η2p = 0.28. This indicates that across age groups making an error in a previous sequence slowed the first key press of the subsequent sequence more than later key presses of that sequence—which however were still significantly slowed (236 ms for key 1 vs. on average 76 ms for keys 2–6). Figure 2 illustrates that this applies to participants in each age group, despite different RT baselines and RT patterns for the three age groups. Results further showed a Trial type x Block interaction, F(1, 60) = 6.27, p < 0. 05, η2p = 0.09, suggesting that the magnitude of post-error slowing differed between the first and second half of the experiment. Although slowing was significant in both halves of the experiment (146 vs. 61 ms), Fs > 19.21, ps < 0.001, η2ps > 0.23, it was larger in the first half. Finally, results of the ANOVA on RTs in the 6-key sequence showed no significant interaction between Trial type and Age group (p = 0.089), indicating that the magnitude of post-error slowing did not differ between the three age groups (see Figure 1, right panel).


Post-error slowing in sequential action: an aging study.

Ruitenberg MF, Abrahamse EL, De Kleine E, Verwey WB - Front Psychol (2014)

Mean RTs per key press within post-error (open circles) and post-correct (black squares) 6-key sequences as a function of age group. Error bars represent standard errors.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3927119&req=5

Figure 2: Mean RTs per key press within post-error (open circles) and post-correct (black squares) 6-key sequences as a function of age group. Error bars represent standard errors.
Mentions: For the 6-key sequence, the data of 23 young adults, 20 middle-aged and 23 elderly participants were included. Results of the ANOVA showed that post-error sequences were generally performed slower than post-correct sequences (714 vs. 612 ms), F(1, 60) = 28.28, p < 0.001, η2p = 0.32. So, the average post-error sequence slowing amounted to 102 ms. An additional Trial type x Key interaction suggested that the amount of post-error slowing differed between the key presses within the sequence, F(5, 300) = 4.00, p < 0.05, η2p = 0.06. Further analyses showed that the interaction was no longer significant after removing the first key press of the sequence from the analysis (p = 0.37), while the main effect of Trial type still remained significant, F(1, 60) = 23.52, p < 0.001, η2p = 0.28. This indicates that across age groups making an error in a previous sequence slowed the first key press of the subsequent sequence more than later key presses of that sequence—which however were still significantly slowed (236 ms for key 1 vs. on average 76 ms for keys 2–6). Figure 2 illustrates that this applies to participants in each age group, despite different RT baselines and RT patterns for the three age groups. Results further showed a Trial type x Block interaction, F(1, 60) = 6.27, p < 0. 05, η2p = 0.09, suggesting that the magnitude of post-error slowing differed between the first and second half of the experiment. Although slowing was significant in both halves of the experiment (146 vs. 61 ms), Fs > 19.21, ps < 0.001, η2ps > 0.23, it was larger in the first half. Finally, results of the ANOVA on RTs in the 6-key sequence showed no significant interaction between Trial type and Age group (p = 0.089), indicating that the magnitude of post-error slowing did not differ between the three age groups (see Figure 1, right panel).

Bottom Line: Importantly, for young adults and middle-aged participants the observed slowing was also accompanied by increased accuracy after an error.We suggest that slowing in these age groups involves both functional and non-functional components, while slowing in elderly participants is non-functional.Moreover, using action sequences (instead of single key-presses) may allow to better track the effects on performance of making an error.

View Article: PubMed Central - PubMed

Affiliation: Department of Cognitive Psychology and Ergonomics, University of Twente Enschede, Netherlands.

ABSTRACT
PREVIOUS STUDIES DEMONSTRATED SIGNIFICANT DIFFERENCES IN THE LEARNING AND PERFORMANCE OF DISCRETE MOVEMENT SEQUENCES ACROSS THE LIFESPAN: Young adults (18-28 years) showed more indications for the development of (implicit) motor chunks and explicit sequence knowledge than middle-aged (55-62 years; Verwey et al., 2011) and elderly participants (75-88 years; Verwey, 2010). Still, even in the absence of indications for motor chunks, the middle-aged and elderly participants showed some performance improvement too. This was attributed to a sequence learning mechanism in which individual reactions are primed by implicit sequential knowledge. The present work further examined sequential movement skill across these age groups. We explored the consequences of making an error on the execution of a subsequent sequence, and investigated whether this is modulated by aging. To that end, we re-analyzed the data from our previous studies. Results demonstrate that sequencing performance is slowed after an error has been made in the previous sequence. Importantly, for young adults and middle-aged participants the observed slowing was also accompanied by increased accuracy after an error. We suggest that slowing in these age groups involves both functional and non-functional components, while slowing in elderly participants is non-functional. Moreover, using action sequences (instead of single key-presses) may allow to better track the effects on performance of making an error.

No MeSH data available.


Related in: MedlinePlus