Short article Intention and attention in ideomotor learning

Transcription

1 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY 2009, 62 (2), Short article Intention and attention in ideomotor learning Arvid Herwig Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany Florian Waszak Laboratoire Psychologie de la Perception, UMR 8158, CNRS and Université Paris Descartes, Paris, France Human actions may be carried out in response to exogenous stimuli (stimulus based) or they may be selected endogenously on the basis of the agent s intentions (intention based). We studied the functional differences between these two types of action during action effect (ideomotor) learning. Participants underwent an acquisition phase, in which each key-press (left/right) triggered a specific tone (low pitch/high pitch) either in a stimulus-based or in an intention-based action mode. Consistent with previous findings, we demonstrate that auditory action effects gain the ability to prime their associated responses in a later test phase only if the actions were selected endogenously during acquisition phase. Furthermore, we show that this difference in ideomotor learning is not due to different attentional demands for stimulus-based and intention-based actions. Our results suggest that ideomotor learning depends on whether or not the action is selected in the intention-based action mode, whereas the amount of attention devoted to the action effect is less important. Keywords: Action-control; Sensorimotor integration; Ideomotor-learning. The host of actions that humans perform every day can be described and analysed in two principal ways. On the one hand, actions are carried out to manipulate the environment on the basis of the agent s intentions or goals (e.g., taking the car to visit a friend). On the other hand, actions are carried out in response to external stimuli to accommodate to environmental demands (e.g., stopping the car at a red traffic light). We refer to these different types of action as intention based and stimulus based, respectively. In the past decades research focused on the differences of intention- and stimulus-based actions regarding the underlying neural mechanisms. These efforts resulted in a large body of evidence suggesting that one and the same overt action can be controlled by different (but partially overlapping) neural substrates, with the medial Correspondence should be addressed to Arvid Herwig, Department of Psychology, Max Planck Institute for Human Cognitive and Brain Sciences, Stephanstraße 1a, Leipzig, Germany. herwig@cbs.mpg.de The research reported here was conducted in partial fulfilment of a PhD thesis by Arvid Herwig. We thank André Spitaler for collecting the data and Joachim Hoffmann and an anonymous reviewer for helpful comments on a previous version of this article. # 2008 The Experimental Psychology Society 219 DOI: /

2 HERWIG AND WASZAK wall of the premotor cortex being primarily involved when the action is intention based and parietal and lateral premotor areas being involved when the action is stimulus based (e.g., Cunnington, Windischberger, Deecke, & Moser, 2002; Goldberg, 1985; Mueller, Brass, Waszak, & Prinz, 2007; Waszak et al., 2005). Although intention-based actions are at the core of humans efficiency to interact with the environment, the majority of studies in the field of psychology investigated stimulus-based actions. One reason for this imbalance might be the difficulty to investigate intentions experimentally. Nevertheless, one approach to investigate intention-based actions by means of experimental psychology can be derived from ideomotor theory (Greenwald, 1970; James, 1890/1950; Prinz, 1997). Ideomotor approaches claim that actions are selected with respect to their perceptual consequences. It is assumed that this ability is acquired in two steps (e.g., Elsner & Hommel, 2001). People first compile associations between movements and their ensuing sensory effects ( action effect bindings ). These associations are bidirectional and can be used in a second step in the reverse direction to voluntarily select an action by anticipating its effect. The validity of the ideomotor principle was supported in several studies (e.g., Elsner & Hommel, 2001; Hommel, Alonso, & Fuentes, 2003; Waszak & Herwig, 2007). For example, Elsner and Hommel made participants first undergo an acquisition phase, in which a self-selected key-press always produced a particular tone (e.g., left key-press! high-pitch tone; right key-press! low-pitch tone). In a second phase, the same tones were used as imperative stimuli for a speeded-choice response. If the key-press was performed in response to the tone that the action had previously produced (e.g., low-pitch tone! right key-press), the response times were faster than they were to a tone that had been previously produced by the alternative action (e.g., high-pitch tone! right key-press). This result indicates that the perception of a learned sensory effect activates the action it is associated with, which can be interpreted as evidence for action effect (or ideomotor) learning. However, recently Herwig, Prinz, and Waszak (2007) have shown that ideomotor learning highly depends on the mode of movement the actions are performed in. In their experiment, ideomotor learning as demonstrated by Elsner and Hommel (2001) occurred only if participants freely selected between left and right key-presses (intentionbased acquisition). In contrast, if the actions that participants performed were triggered by external stimulus events (stimulus-based acquisition) no action effect learning took place whatsoever. To explain this result Herwig et al. suggested that actions are governed with respect to their anticipated sensory consequences only if the agent acts in the intention-based action mode. When acting in the stimulus-based mode, by contrast, participants pass on control to the stimulus (prepared reflex, Hommel, 2000) that is, actions are selected with respect to their antecedents. This difference in action control in turn results in different types of learning: The activity of the system guiding intention-based actions results in action effect or ideomotor learning, whereas the activity of the system controlling stimulus-based actions results in stimulus response or sensorimotor learning. However, note that manipulating the action mode by making participants freely choose between two actions versus responding to target stimuli with one of two actions might entail differences in allocation of attention to the external events involved in the two conditions. In the stimulus-based action mode each action is sandwiched between the imperative stimulus and the effect tone, whereas in the intention-based action mode only the effect tone, but not the preceding stimulus, was presented. As a consequence, participants attention in the stimulus-based action mode might be divided between three elements namely, the imperative stimulus, the response, and the effect tone (S R E). In contrast, in the intention-based action mode, participants might attend only two elements (R E). Assuming that attention is a limited resource and that the strength of each acquired association depends on the attention directed to each element in a given learning situation (Rescorla 220 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (2)

3 INTENTION AND ATTENTION IN IDEOMOTOR LEARNING & Wagner, 1972), the R E associations in the stimulus-based action mode might be weaker than those in the intention-based action mode. If so, then ideomotor learning as demonstrated by Elsner and Hommel (2001) might not depend on the action mode per se, as suggested by Herwig et al. (2007), but merely on the amount of attention devoted to the processing of the effects. This question is pivotal in research on action control. The intention hypothesis put forward by Herwig et al. (2007) is based on the notion that differences in ideomotor learning reflect the fundamentally different, complementary way in which the two routes to action work. The attention hypothesis, by contrast, assumes that they are simply due to the two systems drawing differently on attentional mechanisms. We tested whether ideomotor learning depends on intention or attention in two complementary ways. In Experiment 1, we made the attentional requirements of intention-based actions more similar to the requirements of stimulus-based actions by introducing a stimulus that the participants have to discriminate before making the freechoice action (just as they have to do in the stimulus-based condition). This manipulation was meant to divide participants attention in an intention-based action mode between three elements namely, the preceding stimulus, the response, and the effect (S R E). In Experiment 2, in contrast, we stressed the action effect (R E) relation of stimulus-based actions by directing participants attention towards the effect tone. We did so by presenting catch trials in which the action triggered the wrong effect and which the participants had to detect. Experiment 3 finally controlled whether stressing the R E relation by presenting catch trials leads to ideomotor learning with intention-based actions. If ideomotor learning primarily depends on the attentional resources that are available for the processing of the effect, then transfer effects as demonstrated by Elsner and Hommel (2001) should be observed only in Experiments 2 and 3. If, by contrast, ideomotor learning primarily depends on the mode in which actions are performed during acquisition phase, then one would expect to find transfer effects only in Experiments 1 and 3. EXPERIMENT 1 Method A total of 40 participants (mean age: 24.2 years) took part. The experiment was divided into an acquisition phase and a test phase (see Figure 1 for an overview). Throughout the experiment, a small white cross (þ) presented on a black background in the centre of the screen served as fixation point. The viewing distance was about 70 cm. A trial in the acquisition phase started with the presentation of a small green or red asterisk (mean extension: ), which was displayed about 0.88 below the fixation point. Depending on the colour of the asterisk, participants were instructed either to produce one of two key-presses with their left or right index finger depending on their own choice (go trials), or to omit their response (no-go trials). Each key-press in a go trial directly triggered a particular auditory effect tone (high/low). Participants were not informed about the action effect mapping but were told that tones were presented as a feedback that their responses were recorded by the computer. The action effect mapping was balanced across participants. Auditory stimuli were 200- ms-lasting MIDI tones (instrument oboe) of 392 Hz (low pitch) and 784 Hz (high pitch), presented through the speakers of a headphone. Anticipations, response omissions (go trials), and false alarms (no-go trials) were recorded and fed back by a visual warning message. The next trial started 1,500 ms after the response (go trials) or the no-go stimulus. The acquisition phase comprised 200 go trials and 100 no-go trials. After completing the acquisition phase, participants received a computerized on-screen instruction of the required stimulus response (S R) mapping for the test phase. In each test trial, one of the two effect tones was presented as a target stimulus. There were two subgroups of participants: In the acquisition-compatible subgroup, THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (2) 221

4 HERWIG AND WASZAK Figure 1. Design of Experiments 1 3. Catch trials of Experiment 2 and 3 were presented in 10% of acquisition trials and differed with respect to the effects (E 1 0 and E 2 0) triggered by the response. S ¼ stimulus, R ¼ response, E ¼ effect. participants had to respond with the key that preceded the tone in the acquisition phase. In the acquisition-incompatible subgroup, participants were to respond with the key that preceded the other tone in the acquisition phase. The next trial started 1,000 ms after the response. Participants worked through 200 test trials. Results and discussion The significance criterion was set to p,.05 for all analyses. Violations of sphericity were corrected using the Huynh Feldt 1, and partial h 2 is reported as a measure of the size of an effect. Acquisition phase Participants executed freely selected actions about 386 ms after the onset of the go stimulus. The distribution of left-hand versus right-hand keypresses was nearly equal (49.2 vs. 50.8%). The response was correctly omitted in 97.9% of the no-go trials. There was no significant a priori difference between the acquisition-compatible and acquisition-incompatible subgroups. Test phase Mean reaction times (RTs) and percentages of error were analysed by an analysis of variance (ANOVA) with the between-subjects factor group (acquisition 222 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (2)

5 INTENTION AND ATTENTION IN IDEOMOTOR LEARNING compatible vs. acquisition incompatible) and the within-subjects factor block (40 trials each). As shown in Figure 2 (left panel), the acquisitioncompatible group responded more quickly than the acquisition-incompatible group. The effect of group was the only significant effect, F(1, 38) ¼ 4.55, MSE ¼ 7,612.42, p,.05, h 2 ¼.11, whereas neither the block-factor, F(4, 152) ¼ 1.12, MSE ¼ , 1 ¼.84, h 2 ¼.03, nor the interaction, F,1, h 2 ¼.02, was significant. Response errors were rare (2.9% vs. 2.6% in the acquisitioncompatible and the acquisition-incompatible group, respectively), and the ANOVA of error rates did not produce any effects. Before we interpret this data pattern, however, it is important to show that the implementation of an additional stimulus preceding the action actually captures attentional resources during acquisition. First, it has to be noticed that participants performed the go/no-go task very accurately with only 2.1% responses on no-go trials and only 0.4% response omissions on go trials. Second, in contrast to the RTs of the intentionbased acquisition groups of Elsner and Hommel (2001) and Herwig et al. (2007), which were about ms, response execution times in the present study (386 ms) are comparable to the stimulus-based acquisition groups investigated by Herwig et al. (2007). The data, thus, clearly show that participants actually attended to the go/no-go stimulus and that this manipulation matched the attentional requirements of intention-based actions to those of stimulus-based actions. Importantly, despite this successful manipulation of attention, ideomotor learning took place during the (intention-based) acquisition phase of Experiment 1. EXPERIMENT 2 Experiment 2 was conducted to make the attentional requirements of stimulus-based actions more similar to the requirements of intentionbased actions by guiding participants attention towards the effect tones. Method A total of 40 participants (mean age: 24.8 years) took part. The experiment was divided into an acquisition phase and a test phase (see Figure 1 for an overview). The acquisition phase of Experiment 2 was comparable to that of Experiment 1 with the following exceptions: depending on the colour of the asterisk (red or green), participants were now instructed to respond with either a left or a right key-press (stimulus-based acquisition phase). Each keypress directly triggered a particular auditory effect tone of 392 Hz (low pitch) or 784 Hz (high pitch). In 10% of acquisition trials, the Figure 2. Mean reaction times in the test phase of Experiment 1 (left panel), Experiment 2 (middle panel), and Experiment 3 (left panel) as a function of 40-trial blocks and group. Error bars show standard errors. THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (2) 223

6 HERWIG AND WASZAK effect tone was a semitone below (low pitch: 370 Hz) or above (high pitch: 830 Hz) the standard effect tone. Participants were instructed to respond to these catch trials with a repetition of the previously performed key-press. Data of one participant who did not respond to the catch trials at all were excluded from the analyses. The acquisition phase comprised 234 standard acquisition trials and 26 catch trials. Stimuli and procedures during test phase were identical to those in Experiment 1. Results and discussion Acquisition phase Participants responded to the onset of the asterisk with a mean RT of 423 ms. The hit rate for catch trials was 85.5%, whereas false alarms in standard acquisition trials occurred in 2.7% on average. Importantly, there was no significant a priori difference between the acquisition-compatible and acquisition-incompatible subgroup. Test phase Mean RTs and percentages of error were analysed as in Experiment 1. As shown in Figure 2 (middle panel), the acquisition-compatible group responded as fast as the acquisition-incompatible group. The ANOVA of RTs yielded that neither the block factor, F(4, 148) ¼ 2.15, MSE ¼ 1,304.93, 1 ¼.74, h 2 ¼.06, nor the group factor and the interaction (both Fs,1, h 2 ¼.002 and.02, respectively) was significant. Response errors were rare (2.9% vs. 3.8% in the acquisition-compatible and the acquisition-incompatible group, respectively), and the ANOVA of error rates did not produce any effects. Experiment 2 again confirms that there is no ideomotor learning if participants performed a stimulus-based acquisition phase (Herwig et al., 2007). What is more, Figure 2 indicates that even at the very beginning of the test phase that is, immediately after the acquisition phase no transfer effects are present. This rules out that a weak memory trace simply goes undetected due to fast decay. A post hoc power analyses revealed that large effects ( f ¼.40, which corresponds to h 2 ¼.14) as defined by Cohen (1988) could be detected for the group factor with a probability of 1 b ¼.88, given the sample size (N ¼ 39), an a value of.05, and an average population correlation between the levels of the repeated measures factor of.50. We specified the population effect size as large due to the sample effect sizes obtained in experiments investigating intention-based acquisitions in the present study (Experiments 1 and 3, both fs ¼.35) and previous studies (Elsner & Hommel, 2001; Herwig et al., 2007; d ¼ 0.78 to 1.42; note, d ¼ 0.80 defined by Cohen as large effect). Moreover, the sample size of 216 participants in stimulus-based conditions Herwig and colleagues meta-analysed was sufficient to detect even medium effect sizes ( f ¼.25) with a high power of.96. All power analyses were conducted using G Power 3 (Faul, Erdfelder, Lang, & Buchner, 2007). Most importantly, there is no ideomotor learning even though in Experiment 2 the processing of the effect tones was necessary to correctly report catch trials, suggesting that in the stimulus-based action mode ideomotor learning cannot be boosted simply by making the agent allocate attention to the action effects. To make sure, that the missing compatibility effect is in fact due to the stimulus-based nature of the task and not to some unmeant processes based on the implementation of the detection task, 1 we reran Experiment 2 with an intention-based acquisition phase. 1 Although the task of detecting catch tones among ordinary effects was implemented to guide participants attention toward the effects and, thus, to promote ideomotor learning, it might have worked against ideomotor learning for two unmeant reasons. First, it might be that participants coded the effects merely with regard to their categorical membership (standard tone vs. catch tone) with the consequence that the two standard effect tones were no longer distinguished, and, thus, distinct action effect relations could not be established. Second, it is also possible that participants learned to suppress the additional key-press necessary to indicate a catch tone on the appearance of the standard effect tones, so that the standard effect tone might have served as a no-go signal, and, thus, no compatibility effects were observed in the test phase. We are grateful to Joachim Hoffmann and an anonymous reviewer for pointing this out. 224 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (2)

7 INTENTION AND ATTENTION IN IDEOMOTOR LEARNING EXPERIMENT 3 Method A total of 40 participants (mean age: 24.3 years) took part. Experiment 3 was a replication of Experiment 2 with the only exception that actions in the acquisition phase were freely selected on appearance of a white asterisk (i.e., intention-based acquisition). Results and discussion Acquisition phase Participants executed freely selected actions 329 ms after the onset of the asterisk. Catch trials were correctly detected in 83.6%, whereas false alarms in standard acquisition trials occurred in 2.8% on average. Importantly, there was no significant a priori difference between the acquisition-compatible and acquisition-incompatible subgroups. Test phase The ANOVA of RTs yielded a main effect of group, F(1, 38) ¼ 4.46, MSE ¼ 14,889.60, p,.05, h 2 ¼.11, indicating faster responses in the acquisition-compatible than in the acquisition-incompatible group (309 vs. 345 ms, respectively). This main effect was further modulated by a significant interaction with block, F(4, 152) ¼ 2.65, MSE ¼ 1,132.71, 1 ¼.82, p,.05, h 2 ¼.07. As depicted in Figure 2 (right panel), and as was revealed by t tests, there was a compatibility effect in the first, third, and fourth blocks, t(38) ¼ 2.66, p,.05; t(38) ¼ 2.22, p,.05; t(38) ¼ 2.04, p,.05, but not in the second and fifth blocks, t(38) ¼ 1.34, p ¼.19; t(38) ¼ 1.12, p ¼.27. The main effect of block missed the significance criterion, F(4, 152) ¼ 2.36, MSE ¼ 1,132.71, 1 ¼.82, p ¼.07, h 2 ¼.06. Response errors were rare (2.5% vs. 3.1% in the acquisition-compatible and the acquisition-incompatible group, respectively), and the ANOVA of error rates produced only a main effect of block, F(4, 152) ¼ 5.64, MSE ¼ 1,159.75, p,.001, h 2 ¼.13, indicating a decrease of error rate with progressing test phase. Experiment 3 clearly shows that ideomotor learning occurs, if the actions during acquisition are performed in an intention-based way. Thus, modifying the design by implementing a detection task to guide participants attention toward the effects did not impair ideomotor learning. CONCLUSIONS The main goal of this study was to see whether ideomotor learning depends on the action mode (i.e., intention based vs. stimulus based) or on the attentional resources available for the processing of the effect stimulus. Experiments 1 and 3 showed that auditory action effects gain the ability to prime their associated responses in a later test phase, when the actions performed in the acquisition phase are internally selected. By contrast, Experiment 2 showed that ideomotor learning does not take place when the actions performed in the acquisition phase are exogenously driven. These findings perfectly replicate the selective impact of the action mode on ideomotor learning demonstrated by Herwig et al. (2007). More importantly, guiding participants attention away (Experiment 1) or towards (Experiments 2 and 3) the effect tone, did not influence the pattern of results. Hence, ideomotor learning is boosted whenever the action is selected endogenously, whereas it is hampered whenever the action is selected with respect to some external demand, regardless of the amount of attention devoted to the action effect. Evidently, the detection of transfer effects from an acquisition phase to a test phase depends chiefly on two factors: the strength of the associations formed during acquisition and the sensitivity of the test. The present results show that the strength of the association between actions and their ensuing effects is tremendously affected by the participant s intention, to the point that given the particular test sensitivity of Experiments 1 to 3 a stimulus-based acquisition phase did not result in any observable transfer effect whatsoever. However, ideomotor learning has already been demonstrated for actions that are selected THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (2) 225

8 HERWIG AND WASZAK exogenously (e.g., Elsner & Hommel, 2004; Hoffmann, Sebald, & Stoecker, 2001; Kunde, Hoffmann, & Zellmann, 2002; Ziessler & Nattkemper, 2002). As discussed in detail by Herwig and colleagues (2007) there are two reasons that might account for the divergent results. The first concerns differences in test phase sensitivity between the studies. Research on priming recently demonstrated that the sensitivity to detect priming effects can be enhanced by protracting the operation time of the probe event (Waszak & Hommel, 2007). A number of studies reporting ideomotor learning with stimulus-based actions might have been very sensitive to transfer effects due to their rather slow overall RTs in the test phases (e.g., Elsner & Hommel, 2004; Ziessler & Nattkemper, 2002). The second reason concerns a factor other than intention that might influence the strength of the associations compiled in the acquisition phase: The stimulus-based actions used in the present study were relatively easy or reflex-like compared to all of the studies mentioned above. Using more complex (i.e., less reflex-like) S R mappings might reduce the degree to which participants pass on control to the stimulus so that the action s consequences might become more important in action control. It is thus possible that the difference in results between the intention- and the stimulus-based acquisition phases is rather a question of degree. However, the fact that ideomotor learning might be influenced by several factors not addressed in the present study notwithstanding, our experiments clearly show that differences in ideomotor learning reported by Herwig et al. (2007) are due to the action mode and not to the amount of attention devoted to the processing of the effects. REFERENCES Original manuscript received 8 April 2008 Accepted revision received 12 June 2008 First published online 17 October 2008 Cohen, J. (1988). Statistical power analysis for the social sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Cunnington, R., Windischberger, C., Deecke, L., & Moser, E. (2002). The preparation and execution of self-initiated and externally-triggered movement: A study of event-related fmri. NeuroImage, 15, Elsner, B., & Hommel, B. (2001). Effect anticipation and action control. Journal of Experimental Psychology: Human Perception & Performance, 27, Elsner, B., & Hommel, B. (2004). Contiguity and contingency in the acquisition of action effects. Psychological Research, 68, Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, Goldberg, G. (1985). Supplementary motor area structure and function: Review and hypotheses. Behavioral and Brain Sciences, 8, Greenwald, A. G. (1970). Sensory feedback mechanisms in performance control: With special reference to the ideo-motor mechanism. Psychological Review, 77, Herwig, A., Prinz, W., & Waszak, F. (2007). Two modes of sensorimotor integration in intentionbased and stimulus-based actions. Quarterly Journal of Experimental Psychology, 60, Hoffmann, J., Sebald, A., & Stoecker, C. (2001). Irrelevant response effects improve serial learning in serial reaction time tasks. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, Hommel, B. (2000). The prepared reflex: Automaticity and control in stimulus response translation. In S. Monsell & J. Driver (Eds.), Control of cognitive processes: Attention and performance XVIII (pp ). Cambridge, MA: MIT Press. Hommel, B., Alonso, D., & Fuentes, L. J. (2003). Acquisition and generalization of action effects. Visual Cognition, 10, James, W. (1950). The principles of psychology. Dover Publications, New York. (Original work published 1890.) Kunde, W., Hoffmann, J., & Zellmann, P. (2002). The impact of anticipated action effects on action planning. Acta Psychologica, 109, Mueller, V., Brass, M., Waszak, F., & Prinz, W. (2007). The role of the presma and the rostral cingulate zone in internally selected actions. NeuroImage, 37, Prinz, W. (1997). Perception and action planning. European Journal of Cognitive Psychology, 9, THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (2)

9 INTENTION AND ATTENTION IN IDEOMOTOR LEARNING Rescorla, R. A., & Wagner, A. R. (1972). A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and non-reinforcement. In A. H. Black & W. F. Prokasy (Eds.), Classical conditioning II: Current research and theory (pp ). New York: Appleton-Century-Crofts. Waszak, F., & Herwig, A. (2007). Effect anticipation modulates deviance processing in the brain. Brain Research, 1183, Waszak, F., & Hommel, B. (2007). The costs and benefits of cross-task priming. Memory and Cognition, 35, Waszak, F., Wascher, E., Keller, P., Koch, I., Aschersleben, G., Rosenbaum, D., et al. (2005). Intention-based and stimulus-based mechanisms in action selection. Experimental Brain Research, 162, Ziessler, M., & Nattkemper, D. (2002). Effect anticipation in action planning: Anticipative learning of action effects. In W. Prinz & B. Hommel (Eds.), Common mechanisms in perception and action: Attention and performance XIX (pp ). Oxford, UK: Oxford University Press. THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2009, 62 (2) 227