The limits of parallel processing [Elektronische Ressource] / Katrin Göthe

Katrin Göthe

Sujets

• Psychologie

Informations

• Publié par: universitat_potsdam
• Publié le: 01 janvier 2009
• Nombre de lectures: 12
• Langue: English

Extrait

The limits of parallel processing

Katrin Göthe
Allgemeine Psychologie I
Universität Potsdam




Dissertationsschrift
zur Erlangung des Grades des Doktors der Philosophie (Dr. phil.)
im Fach Psychologie





Eingereicht bei der
Humanwissenschaftlichen Fakultät
der Universität Potsdam
im Jahr 2009 This work is licensed under a Creative Commons License:
Attribution - Noncommercial - No Derivative Works 3.0 Germany
To view a copy of this license visit
http://creativecommons.org/licenses/by-nc-nd/3.0/de/deed.en






































st1 reviewer: Prof. Klaus Oberauer
nd2 reviewer: Prof. Reinhold Kliegl
Day of oral defense: 15. December 2009

Published online at the
Institutional Repository of the University of Potsdam:
URL http://opus.kobv.de/ubp/volltexte/2010/4606/
URN urn:nbn:de:kobv:517-opus-46063
http://nbn-resolving.org/urn:nbn:de:kobv:517-opus-46063 Acknowledgements

This work could not be realized unassisted. I am deeply grateful for the support of
many people and decided that saying less is more:

Thank you!





Table of Contents
SUMMARY............................................................................................................................... 7
1. INTRODUCTION.......................................................................................................... 11
1.1. The present thesis ..................................................................................................................................12
1.2. The Bottleneck theory ...........................................................................................................................13
1.3. Resource sharing accounts....................................................................................................................15
1.4. The Executive-process interactive-control model ...............................................................................16
2. EXPERIMENT 1 -DUAL-TASK PROCESSING OF TWO MEMORY UPDATING
TASKS WITH ARBITRARY S-R MAPPINGS ................................................................. 20
2.1. Method....................................................................................................................................................26
2.2. Results.....................................................................................................................................................29
2.3. Discussion ...............................................................................................................................................40
3. EXPERIMENT 2 – TWO TASKS WITH NON-COMPATIBLE S-R MAPPING
VS. TWO TASKS WITH ONE TASK CONTAINING A COMPATIBLE S-R
MAPPING............................................................................................................................... 46
3.1. Method....................................................................................................................................................48
3.2. Results.....................................................................................................................................................49
3.3. Discussion ...............................................................................................................................................55
4. EXPERIMENT 3 –DUAL-TASK PROCESSING OF S-R AND C-R MODALITY-
PAIRING GROUPS OVER PRACTICE WITHIN A SPP................................................ 58
4.1. Method....................................................................................................................................................64
4.2. Results.....................................................................................................................................................67
4.3. Discussion ...............................................................................................................................................77
5. EXPERIMENT 4- DUAL-TASK PROCESSING OF S-R AND C-R MODALITY-
PAIRING GROUPS AFTER PRACTICE WITHIN A PRP PARADIGM...................... 87
5.1. Method....................................................................................................................................................94
5.2. Results.....................................................................................................................................................95
5.3. Discussion .............................................................................................................................................112
6. GENERAL DISCUSSION........................................................................................... 117
5 6.1. An integrated view of dual-task costs ................................................................................................117
6.1.1. S-R compatibility..............................................................................................................................117
6.1.2. Modality pairings..............................................................................................................................117
6.1.3. Practice .............................................................................................................................................119
6.1.4. Interindividual differences................................................................................................................120
6.1.5. The response selection bottleneck ....................................................................................................121
6.2. Response selection revised: An alternative information processing account .................................123
6.3. Conclusion............................................................................................................................................126
7. LITERATURE ............................................................................................................. 128
8. APPENDICES .............................................................................................................. 133
8.1. Appendix A: Serial processors of Experiment 2. ..............................................................................133
8.2. Appendix B: Analyses of the accuracy data for Experiment 3 ........................................................135
8.3. Appendix C: Analyses of the single-task trials of Experiment 4 .....................................................142



6 Summary
Doing two things at once should be the fastest and simplest solution for all people that
want to manage more in less time. However, there seem to be severe limits for parallel
processing. Mostly, when one performs two tasks at the same time, performance of one or
both tasks decreases compared to the situation when one performs each task by itself (Pashler,
1984; Ruthruff, Hazeltine, & Remington, 2006; Ruthruff, Johnston, Van Selst, Whitsell, &
Remington, 2003; Van Selst & Jolicoeur, 1994). The present thesis deals with the question to
which extent people are able to process two cognitive tasks at the same time and why costs
emerge in the majority of cases. Among researchers a strong controversy exists about the
limitations of our cognitive system. Generally dual-task interference is taken as evidence for a
capacity limitation of our information processing system.
One dominating theoretical model explains interference with a processing bottleneck
(Pashler, 1994a). This bottleneck operates at the central response selection stage. This stage is
intermediate between stimulus encoding and response execution (Sternberg, 1969). The
response selection bottleneck allows only one task at a time to be processed centrally. Is the
bottleneck occupied with one task, the next task that arrives at the bottleneck has to wait until
it is released by Task 1. This waiting time increases overall processing duration of Task 2
compared to a single-task situation. According to this model stages before and after the
bottleneck can be processed in parallel between the two tasks. Therefore, dual-task costs that
come up on Task 2 should depend on the duration of the central processing time of Task 1.
Moreover, parallel processing is predicted to be not possible. Dual-task costs should be
omnipresent whenever the central processing stages of two tasks had to be processed
simultaneously.
The bottleneck model has faced some challenges during the last years, because for
some task combinations vanished dual-task costs were revealed (Hazeltine, Teague, & Ivry,
2002; Oberauer & Kliegl, 2004; Schumacher et al., 2001) implying that parallel processing is
possible. Taking two of these as a starting point the present thesis focuses on factors that are
likely to promote parallel processing. For that purpose the thesis examines whether costs
reemerge when the similarity between stimulus (S) and response (R) representations of one
task is reduced (Expe