Effects of absolute and relative practice on n − 2 repetition costs
Introduction
In everyday life, humans are confronted with the need to continuously adapt their behavior in accordance with changes in the environment. To investigate this flexible control of behavior experimentally, the task switching paradigm is a widely used tool: Participants have to switch between two or more simple classification tasks. As a result, reaction times (RT) and error rates (ER) are higher when the task switches compared to repetition trials, an effect termed switch cost or shift cost (see Kiesel et al., 2010, Vandierendonck et al., 2010, for reviews). Two types of cognitive control processes are commonly assumed to be involved in task switching, namely the activation of the currently relevant task on the one hand and the inhibition of competing information on the other hand.
In task switching research, n − 2 repetition costs (also labeled backward inhibition effect, Mayr & Keele, 2000) are one of the most straightforward indications for an involvement of inhibitory processes. N − 2 repetition costs are indicated by the observation that, when switching among three tasks A, B, and C, reaction times (and sometimes also error rates) are higher for task sequences ABA compared to sequences of type CBA. That is, switching back to a recently abandoned task is accompanied by costs that are taken as evidence for the need to overcome inhibition attached to that task (cf. Koch, Gade, Schuch, & Philipp, 2010, for a review). However, although there are many studies showing the occurrence of n − 2 repetition costs in different experimental settings, many aspects of this effect still remain unclear.
Recently, Grange and Juvina (2015) showed that n − 2 repetition costs decline with increasing practice. This effect was explained by the automization of cue–task translation processes. Referring to Logan (1988), cue–task1 translation processes are slow at the beginning of the experiment, because the cue indicating the identity of the upcoming task was formerly unrelated to the task (at least when using nontransparent cues). With increasing practice, cue–task associations accumulate in long term memory, causing a more automatic and, therefore, faster retrieval. On a complementary level, Grange and Juvina explained their results by a computational model of inhibition implemented in ACT-R (Grange, Juvina, and Houghton, 2013). For task performance, the successful retrieval of chunks of information from memory is needed, with the speed of retrieval being positively related to the activation level of the chunk. The extent of activation depends on the current context and on base-level learning, which represents the amount of practice with that chunk. Additionally, an inhibition parameter is implemented that is subtracted from base-level learning. As practice increases, base-level learning gets stronger until it outweighs inhibition, which then results in decreasing n − 2 repetition costs. To fully account for the data pattern, Grange and Juvina extended this model, including the assumption of increasing cue–task association strength and decreasing activation noise with practice.
However, based on the experimental design of Grange and Juvina (2015) it is not possible to distinguish whether practice in absolute terms or the relative strength of the tasks (i.e., the activation of one task in relation to the activation of the other tasks) governed the reduction in n − 2 repetition costs with time, because all tasks were practiced to the same degree. That relative task strength can indeed have an influence on n − 2 repetition costs was shown by Scheil and Kleinsorge (2014). In their study, preparation time was varied in a trialwise fashion to investigate the influence of preparation time of trials n − 2, n − 1, and n on n − 2 repetition costs. N − 2 repetition costs were highest when preparation time in both trial n − 2 and n − 1 was long. It was argued that a high amount of activation of the tasks to be inhibited (due to sufficient preparation in trial n − 2) as well as a high amount of activation of the alternative task (due to sufficient preparation time in trial n − 1) causes high n − 2 repetition costs, because the necessity of inhibition as well as the possibility for it are high.
To further elucidate this issue, a task switching experiment with three tasks was designed in which one of the tasks was slightly changed after half of the blocks to decrease the relative strength of this particular task. Because the theoretical frameworks Grange and Juvina used to explain the practice effect on n − 2 repetition costs involved cue-related processes as a main aspect, it seems straightforward to aim at cue-related task components to vary the relative task strength. However, cue-related processes as the locus of inhibition cannot be taken for granted as there are also findings suggesting different targets. For example, some authors argue that inhibitory processes causing n − 2 repetition costs are response-related (e.g., Schuch & Koch, 2003). Furthermore, inhibition might be attached in a flexible way to the aspect of the task set causing greatest inter-trial conflict (Houghton, Pritchard, & Grange, 2009). Therefore, two experiments were conducted. In one experiment, one of the stimulus–response mappings was reversed whereas in the second, the cue of one task was changed. Consequently, the aims of the present study were twofold. First, a differentiation between effects of absolute practice and relative strength of the tasks on n − 2 repetition costs was intended. Second, with varying different aspects of the tasks, response-related or cue-related, further insight into the target of inhibitory processes in task switching should be gained. If absolute practice guides the reduction of n − 2 repetition costs, they should become smaller during the course of the experiment without being influenced by changing one of the tasks. If relative task strength induces the reduction, the change of one of the tasks should affect n − 2 repetition costs. In this respect, an effect of the task change in the first experiment would point towards a response-related locus of inhibitory processes, whereas an effect in the second experiment would be indicative for cue-related processes being the target of inhibition. If changing one task influences n − 2 repetition costs in both experiments, a flexible target, like the aspect causing greatest inter-trial conflict, would be a possible explanation.
Section snippets
Participants
24 subjects (6 male) with normal or corrected-to-normal vision participated. Their mean age was 22.56 years (range: 18–30). Participants were assigned alternately to one of the three groups according the order of their appearance.
Stimuli, tasks, and apparatus
Stimuli consisted of two different shapes (x and +) presented in yellow or blue and with a size of either 3 cm × 3 cm or 6 cm × 6 cm. Task cues consisted of a dark gray diamond, square, or triangle surrounding the position of the imperative stimulus with a size of about 7 cm × 7
Participants
24 subjects participated. One participant had to be excluded due to an ER of > 38%. The final sample consisted of 7 men and 16 women with a mean age of 24.17 years (range: 19–30). All had normal or corrected-to-normal vision. Participants were assigned to one of the three groups according the order of their appearance.
Stimuli, tasks, and apparatus
These were the same as in Experiment I.
Design and procedure
These were the same as in Experiment I with the difference that response mappings were left unchanged throughout the experiment, but instead
General discussion
The present study aimed at dissociating effects of absolute and relative task strength on n − 2 repetition costs. While absolute practice was varied by time on task, the relative strength was varied via changing aspects of one of the three tasks during the course of the experiment. For this purpose, the stimulus–response mapping of one task was changed in Experiment I and the cue of one task in Experiment II. In Experiment I, decreasing n − 2 repetition costs with increasing practice occurred, as
Ethical standards
The procedures employed in the present study were approved by the ethics committee of the Leibniz Research Centre for Working Environment and Human Factors.
Conflict of interest
The author declares that she has no conflict of interest.
Acknowledgments
I thank Thomas Kleinsorge for helpful discussions and Florian Selchow for running the experiments.
References (15)
- et al.
The effect of practice on n − 2 repetition costs in set switching
Acta Psychologica
(2015) Cue-independent task-specific representations in task switching: Evidence from backward inhibition
Journal of Experimental Psychology: Learning, Memory, and Cognition
(2007)- et al.
Highly proficient bilinguals implement inhibition: Evidence from n − 2 language repetition costs
Journal of Experimental Psychology: Learning, Memory, and Cognition
(2015) - et al.
Dissociating cue-related and task-related processes in task inhibition: Evidence from using a 2:1 cue-to-task mapping
Canadian Journal of Experimental Psychology
(2008) - et al.
Heightened conflict in cue-target translation increases backward inhibition in set switching
Journal of Experimental Psychology: Learning, Memory, and Cognition
(2010) - et al.
On costs and benefits of n − 2 repetitions in task switching: Towards a behavioural marker of cognitive inhibition
Psychological Research
(2013) - et al.
The role of cue-target translation in backward inhibition of attentional set
Journal of Experimental Psychology: Learning, Memory, and Cognition
(2009)