54 research outputs found
Terminating continuous movements with(out) perceptual overlap to moving
No full textThis experiment tests the idea of ‘termination by a new plan’ for continuous movements. Also, it studies body-external perceptual features of termination (e.g., Mocke et al., 2020). It should be harder to stop an action (A) when it shares features with the perceptual consequences of the ‘new act’ to not move (B) than when it does not, because the new plan to not move would retrieve features of the plan to move. This experiment is the first demonstration of this, as we believe, novel phenomenon.
Participants move their hand on a digitizer tablet either clockwise or counterclockwise (varied randomly across trials) in a certain speed. At an unpredictable time point a visual halt signal is presented (color change of fixation cross), which requires to immediately terminate that movement. Minimum movement duration will be 3 seconds. These 3 seconds will be added to an additional duration, which we will sample from a nonaging (e.g., exponential) distribution with rate lambda = 1 (that is, with mean 1) to countermand aging foreperiod effects.
Doing so causes in different blocks (with block order counterbalanced across participants), and thus foreseeably, a visual movement that either does or does not share the direction feature (clockwise or counterclockwise rotation) with the to be terminated hand movement.
One group will be instructed to stop their movement as soon as the stimulus appears, while the other group will be instructed to initiate a small counter-movement (a few millimeters large )
Contrast & Assimilation in Action Sequences (ABBA task)
No full textPrevious research has shown that action durations can be influenced by action plans held in working memory for later (Mocke et al., 2022). Specifically, keypresses tend to be shorter when a long action is planned than when a short action is planned. This hints towards a contrasting mechanism between planned and first-to-be-executed actions. In previous studies, actions were additionally defined by their location, not only their duration. In this experiment, we will test whether this contrast effect also occurs for simpler actions, that is, keypresses that are only distinguished by their duration. Further, we will use three different durations and a no plan condition to get an even better understanding of the mechanisms underlying this contrast effect.
In an ABBA paradigm, participants will only use one key. In each trial, they will first see a fixation cross of 300ms, then for 500ms a stimulus A (1, 2, 3 or X) that tells them whether and which duration to plan. After a blank of 1000ms, they will see a 200ms stimulus B (K, M, L) that tells them duration B, and are asked to execute first action B and then A, within a time window of 4000ms after stimulus B onset. Short durations will be between 175 and 350, medium ones between 359 and 700 and long ones between 700 and 1400 ms
Binding of Colour Features and Spatial Features
No full textBased on ideomotor theory, the Theory of Event Coding (Hommel, Müsseler, Aschersleben, & Prinz, 2001) construes action planning as activating and binding features of their anticipated effects. Previous research has shown worse performance for actions that partially overlap with an action plan held in preparation regarding bound action effect features than for actions which features do either not or completely overlap with an active action plan (Mocke, Weller, Frings, Rothermund, & Kunde, 2020).
This online experiment aims at testing whether coding the very same actions in spatial features can be overcome by action coding in non-spatial (color) features. Finding that such recoding is possible would suggest that spatial features are not special but can be replaced by other features optionally.
Participants position one hand "above" (keys 6 and 7) the other hand (keys B and N) on a QWERTZ keyboard. They see differently colored squares aligned in a square around the fixation cross on screen, each representing one of the four used fingers (little and ring finger of one hand, middle and index finger of the other hand). In one of the between-subjects conditions, spatial ("left" and "right") and color ("blue" and "yellow") features will be redundant between the upper and the lower hands, while in the other condition, colors will be reversed for the lower hand.
Actions are instructed in terms of colors and hands. A first stimulus for instance tells participants to prepare a “blue” or “yellow”, "upper" or "lower" keypress as action A. Then, a second stimulus requires a “blue” or “yellow”, "upper" or "lower" keypress as action B. To test for potential spatial coding we will analyze performance in action B in terms of overlap with the spatial features “left” and “right”
Cancellation of Planned Actions: Behavioural measures and EEG correlates (Exp. 2)
No full textDiscarding an action can occur because an initial plan can no longer be carried out. What happens with action files after such discarding? In this experiment, we will investigate this using both behavioral measures and EEG correlates of action plans.
We will modify the standard ABBA paradigm (Mocke et al., 2022), in which participants are asked to prepare an action A. Yet, before carrying out that action another action B is requested. Initiating this first requested action B is typically delayed and more error-prone if it does share some features with the planned action A, than when it does not, leading to so-called partial overlap costs (e.g., Stoet & Hommel, 1999; Wiediger & Fournier, 2008). Usually, performance is best when actions A and B fully overlap. Between stimulus A and stimulus B, we will include a signal indicating that the plan for action A is to be retained for later (75%) or can be discarded (25%). The crucial comparison relates to performance in action B in “discard” and “continue” trials as a function of overlap with the planned or discarded action A (i.e., partial overlap costs).
A combination of word stimuli (DIT for short and DAH for long) with asterisks to the left or right, and of short or long arrows pointing to the left or right, will indicate whether a left or right action of either short or long duration is required.
Throughout each trial, the center of the screen will be framed by a rectangle, that is initially grey. After a jittered fixation cross, participants will see stimulus A (500ms), followed by the planning interval (2000ms). After that, the frame will change to one of two colours (yellow vs. blue), indicating that the plan for action A is to be retained (75%) or discarded (25%). After a certain discard time (1000 ms), stimulus B appears (200ms), and responses B and A are required within another 4000ms after stimulus B onset.
In this second experiment, we increased the discard time from 400 to 1000ms in order to be better able to find EEG markers of discarding without influences of stimulus B and action B
Cancellation of Planned Actions: Manipulation of Discard Time
No full textDiscarding an action can occur because an initial plan can no longer be carried out. What happens with action files after such discarding?
We will modify the standard ABBA paradigm (Mocke et al., 2021), in which participants are asked to prepare an action A. Yet, before carrying out that action another action B is requested. Initiating this first requested action B is typically delayed and more error-prone if it does share some features with the planned action A, than when it does not, leading to so-called partial overlap costs (Stoet & Hommel, 1999; Wiediger & Fournier, 2008). Usually, performance is best when actions A and B fully overlap. Between stimulus A and stimulus B, we will include a signal indicating that the plan for action A is to be retained for later (75%) or can be discarded (25%). The crucial comparison relates to performance in action B in “discard” and “continue” trials as a function of overlap with the planned or discarded action A (i.e., partial overlap costs).
How a discarded action A impacts the initiation of a feature-overlapping action B conceivably depends on two factors, the time already spent on planning action A before discarding, and the time after discarding action A, but before the requested action B. These influences can be distinguished by manipulating the time intervals between the announcement of action A and discard signal, and between discard signal and stimulus B. In this experiment, we will vary the Discard Time trialwise (200ms vs. 400ms).
A combination of DIT/DAH stimuli with asterisks to the left or right (Action A), and of short or long arrows pointing to the left or right A (Action B), will indicate whether a left or right action of either short or long duration is required.
Throughout each trial, the center of the screen will be framed by a rectangle, that is initially grey. After a fixation cross (300ms), participants will see stimulus A (500ms), followed by the planning interval (2000ms). After that, the frame will change to one of two colours (yellow vs. blue), indicating that the plan for action A is to be retained (75%) or discarded (25%). After a certain discard time (200ms vs. 400 ms), stimulus B appears (200ms), and responses B and A are required within another 4000ms after stimulus B onset
Altering bindings of action effect features by affective enrichment 2
No full textPrevious work has already reported preliminary evidence for integration of affective features in event-files (Eder, Müsseler, & Hommel, 2012). Closely linked is the question whether assigning a certain valence to an action effect results in increased relevance and therefore altered binding effects (see Mocke et al., 2020). One first indicator for such an influence of action effect valence on binding is a study showing that rewarding actions can strengthen action-effect bindings (Eder, Erle, & Kunde, 2019). Here, participants learned in an acquisition phase that certain actions (i.e., button presses) result in auditory effects while some but not all actions are financially rewarded. In a test phase, they found prime compatibility effects in RTs to be larger for previously rewarded keypresses in comparison to previously unrewarded or punished keypresses.
Instead of such long-term bindings, one can also investigate whether trialwise formed episodic response-effect (R-E) bindings can be strengthened by reward. Specifically, we will examine this in the context of action planning (instead of action execution). We will test whether when planning an action, which will foreseeably have a particular environmental effect, the strength of the R-E binding (and thus of the not yet executed action plan) can be altered by rendering the effect rewarding, neutral or punishing.
To do so, we will adapt the Inducer-Diagnostic (ID) paradigm by Theeuwes, De Houwer, Eder, and Liefooghe (2015). Here, participants learned two R-E associations via instructions in each run (e.g., 'If you press left, "P" appears. If you press right, "Q" appears). While maintaining these two action plans and before executing them at the end of the run, the subjects worked on a different diagnostic task. The presence of the initially instructed R-E bindings was reflected in a congruency effect, which was caused by improved (reduced) performance if the required action in the intervening trials did (not) correspond to the action linked to the letter stimulus by instruction.
This second experiment will consist of three blocks (positive, neutral, negative). In each, the two environmental effects per run will be assigned the same valence specific for that block. Valence will be operationalised by financial reward/loss (+/- 50p). We will examine whether the valence of the instructed action effects changes the size of the congruency effect.
Compared to the first experiment, the design will be simplified with congruency effect as dependent variable instead of performance itself. Importantly, as the previous experiment did not yield overall congruency effects, we will use an improved paradigm, which will also more closely resemble the original study by Theeuwes and colleagues (2015)
Me or you? Binding and retrieval of multiple response features when representing own and others’ action plans
No full textStoet and Hommel (1999) found that when a person plans an action, essential features of that action are temporarily bound together in so called “action files” (see also Hommel et al., 2001). Further research on this topic showed that binding also occurs not only when spatial features are bound together, but also when a person plans an action that consists of spatial and non-spatial features. This was investigated, for example, with temporal (short and long) and spatial (left and right) response features (Mocke et al., 2022). Having found solid evidence for feature binding when one person plans an action, the question arises whether binding of essential action features occurs also when that person observes another person planning an action. Giesen et al. (2014) have already investigated a similar question and found that binding of stimulus and response features also occurs when one person merely observes another person executing an action. Taking this information together, we want to further examine how binding of two response features (color and duration) for representing an action plan is influenced by the person who is planning the action (oneself or another person). In other words, we will investigate whether the structure of the mental representations of other people’s action plans resembles the structure of our own action plans. To investigate this question, we will use a modified version of the ABBA paradigm. The standard ABBA paradigm consists of a presented stimulus A containing information on the basis of which a person should plan to perform action A (Stoet & Hommel, 1999). However, before the person can carry out action A, they are asked to plan another action B after stimulus B has been presented to them, and to execute that action B prior to action A. In this version, actions A and B will consist of either a long or short button press with the left or right hand. The execution of the first required action B is typically delayed and more error-prone if action A and B have one feature in common compared to if either all or none of the features of action B correspond with action A. This phenomenon is known as “Partial Overlap Costs”. In this study we will use the ABBA paradigm by Mocke et al. (2022) but with a few modifications. First, we will have two subjects sitting across from each other at a table so that they can perform the trials together. The subjects will be instructed to work together as a team gaining 10 points for correctly completed trials and losing 10 points for incorrectly completing any action in the trial. This will create a positive interdependence between the two subjects which should enhance episodic binding effects (see Giesen et al., 2014). Stimuli will be presented and responses given via a tablet that will be placed in the middle of the table at the same distance from both subjects.
Six circular buttons will be presented on the tablet at all times. Four neutrally colored starting areas, one in each of the tablet’s corners, and two target areas (cyan and magenta). Each participant will need to rest their index fingers on the two starting areas closest to them. Correct responses will consist of the correct participant releasing the respective starting area and making a reaching movement to the correct target area (with the hand on the same side as the area), pressing the target area for the correct duration, and moving the index finger back into the starting area. Stimuli will be presented twice in the center of the screen, one being mirrored, so that each can be read by one participant.
We will manipulate the essential features of the action, that is, apart from spatial (here: color) and temporal features also which person executes the action. For action B, this can be subject 1 or subject 2. For action A however, it means either subject 1, subject 2, or neither of the subjects plans to execute action A. In the latter control condition, no subject will be instructed to plan executing action A. This is implemented by a stimulus A that is presented to both subjects, but is missing the information on the person who is to plan action A. The presented stimulus will still contain the relevant color and duration features, but neither subject will be asked to plan an action. We implemented this control condition to test the representation of other people’s actions not only against own actions but also against the representation of a stimulus that does not require any response.
We will also manipulate the duration and color features. The duration feature refers to whether a short (0-200ms) or a long button press (201-700ms) will be required for each action. The color feature refers to whether a press on the cyan or the magenta colored button is required. Since the button that is the ‘left’ button for one participant is automatically the ‘right’ button for the other participant, we will use colored buttons instead of labeling them as “left” and “right”, as in previous experiments. Taken together, the actions consist of one person pressing either the cyan or the magenta key for a short or long period of time.
Having explained the manipulation of our essential action features, it is of course important to define which stimuli will be used in our experiment. The duration feature of Stimulus A consists of either a “DIT” (implying a short key press) or a “DAH” (implying a long key press). We will signal the color feature of action A by using different colors for the word “DIT” versus “DAH”. The word will be colored in cyan or magenta to indicate which button should be pressed. The person feature of stimulus A will be an underlining of the word “DIT” or “DAH” (this subject is being asked to plan action A) while the other subject will see the same word with a line above the word (this subject is not planning action A). If the word is not underlined at all (for both subjects), no one is asked to plan action A (i.e. the control condition is at work).
It is important that stimulus B differs from stimulus A in its essential characteristics. For this reason, we will use an arrow to indicate which colored button should be pressed (color feature). The arrow points to the left if the key is to be pressed on the left and to the right if the key is to be pressed on the right. The buttons will still be colored cyan and magenta. The duration feature of stimulus B will be implemented using the length of the arrow. We will use either a long or a short arrow to indicate whether a short or a long key press is required. We will manipulate the person feature by writing “1” or “2” in the arrow for each subject to indicate whose turn it is.
The exact procedure of one trial is as follows: After a fixation cross (300ms), subjects are presented with stimulus A (500ms), followed by the planning interval (2000ms). Subsequently they will see stimulus B (200ms) and response B will be required within a further 3000ms of stimulus B onset and will be followed by response A within another 3000ms (if it is not a no plan trial). There will be a total of 384 trials, divided into 4 blocks of 96 trials each
How do we represent other people’s action plans?
No full textWe are investigating the differences between cognitive representations of other people’s and own action plans. It seems when planning an action, essential action features are temporarily bound together into so called “action files” (Hommel et al., 2001). Are the same cognitive processes at work when representing an action plan of another person as when planning one’s own action?
We will modify the standard ABBA paradigm (Mocke et al., 2021), in which participants are asked to prepare an action A. Yet, before carrying out that action another action B is requested. Actions can be left or right, short or long keypresses. Initiating this first requested action B is typically delayed and more error-prone if it does share some features with the planned action A, than when it does share neither or all of its features, leading to so-called partial overlap costs (e.g., Stoet & Hommel, 1999; Wiediger & Fournier, 2008).
We will use this paradigm with a few changes. Firstly, there will be two participants sitting next to each other in each experiment executing the trials together. Secondly, the essential features manipulated will be which person is supposed to execute the action (personal feature, equivalent to left/right features since one person is sitting on the left, one on the right) and the duration of the action (temporal feature). Therefore, actions will consist of simple presses of one key per person on a keyboard for either short or long duration.
Thirdly, we will include a control condition in which stimulus A contains information about the temporal but not personal feature (and means no participant has to plan action A) and therefore should neither lead to a representation of an own action nor that of another person. This is in order to test the representation of other people’s actions not only against the representations of own actions but also against the representation of a stimulus that does not require any response.
In order to make the experiment more fun for participants and closer to real life experience participants take on the identity of different opera singers, one male and one female, in a concert setting and follow the instructions of a “conductor” (B) and reading the notes of the music piece (stimulus A). The duration of the keypress models holding a tone for a long vs. short period of time, the horizontal or person feature models the pitch of the voice or sex of the singer.
Stimulus A therefore can either show a note with a tail downwards (high pitch/right participant) or upwards (low pitch/left participant) and filled in body (short) or not (long). And stimulus B can either show the picture of a conductor’s hand raising the baton high (long) or low (short) and to the left (low pitch/left participant) or right (high pitch/right participant).
The exact process of one trial will be as follows: After a fixation cross (300ms), participants will see stimulus A (500ms), followed by the planning interval (2000ms). After that, stimulus B appears (200ms), and responses B and A are required within another 4000ms after stimulus B onset. Each key release after a correctly executed action will produce an auditory effect of the respective (i.e., short vs. long) duration in high or low pitch, depending on the actor. Another practical implication of using sound in our experiment is the facilitation of the execution of action A in the case of different participants conducting action B and A.
There will be 336 trials in total (48 trials in each of the experimental 7 blocks), resulting in 28 trials per experimental cell per person
Cancellation of Instructions in Short and Long Term Memory
No full textBased on the widespread assumption of a capacity-limited working memory, this study examines its ability to discard no longer relevant instructions. Previous studies targeting this issue have led to different results. While many support the existence of inhibitory processes (e.g., seen in task switching costs, Vandierendonck et al., 2010) or n-2 task repetition costs (Schuch & Koch, 2003) or even successful action plan discarding (Mocke et al., 2023), Abrahamse et al. (2023) conducted several experiments using the Inducer-Diagnostic Paradigm (Liefooghe et al., 2012) and were unable to find the same. This study aims to test whether long term memory traces may have played a role in the unsuccessful instruction discarding in Abrahamse et al.’s (2023) experiments, therefore mostly replicating one of them (Experiment 3) except for changes in the Inducer task.
In the Inducer-Diagnostic Paradigm (Liefooghe et al., 2012), two stimulus-response (S-R) mappings (for the Inducer and the Diagnostic Task, respectively) need to be kept in working memory at once. Typically, in the beginning of each run, participants have to memorize two Inducer S-R mappings (e.g., tree-left and doll-right). Before applying those, participants need to respond to the format of the same word stimuli (italics vs. upright) using the Diagnostic S-R mapping instructed in the beginning of the experiment. This results in congruent (if word and format suggest the same response) and incongruent trials (if they suggest a different response). Trials containing incongruent S-R connections should be more difficult than congruent trials since the incorrect response is interfering. Therefore, if subjects can successfully discard the Inducer Task Instruction upon a cue, RTs should not differ anymore between congruent and incongruent trials, since the Inducer Task Instruction should be deleted from working memory and not interfere any longer. To examine this like Abrahamse et al. (2023), there will be congruent and incongruent trials across three different Run Types (Complete vs. Proceed vs. Cancel), in which subjects cancel or complete the Inducer task before the Diagnostic Task trials, whose RTs will be measured. If they only complete the Inducer Task after this set of Diagnostic Task trials, subjects will probably keep the Inducer Task instruction present in working memory during the whole run.
Abrahamse et al. (2023) used the same S-R mapping for the Diagnostic Task but introduced new Inducer Task S-R mappings and new stimuli (i.e., new word pairs) for each run. This study, however, will be using the same two words throughout the experiment, only reassigning left/right responses randomly at the beginning of each run.
Since Abrahamse et al. (2023) never reused the words across runs, a fixed response was assigned to each word, possibly forming a long term memory trace. Perhaps, RT differences between incongruent and congruent trials in the Cancel and Complete conditions were not caused by unsuccessful dismantling on working memory level, but instead by long term memory traces. Contrary to this, the present study aims to prevent this by constantly reassigning the responses to the same two words. Hence, there is no fixed or meaningful connection between a word and a keypress direction since it is changing all the time
Does an Event File Persist after Rebinding one of its Response Features?
No full textThis experiment is based on the Theory of Event Coding (Hommel, Müsseler, Aschersleben & Prinz, 2001) and investigates whether a feature can be bound/stored in only one event file at a given time or in several event files. The Theory of Event Coding states that action and perception share a form of representation (a "common code"). In line with ideomotor theory, it thus claims that actions are represented by features of their perceptual consequences. These features of produced events (body-related or environmental action effects), together with features of perceived events (stimuli) are first activated and then bound to form a so-called "event file".
For instance, in distractor-response-tasks, an event file should contain not only features relating to the response (or target) but also to the irrelevant distractor. This has been shown many times, as responses in a probe trial are slower and more error-prone when the distractor or the response feature overlapped between probe event file and a previously formed prime event file than when both features or none of the features overlapped (Hommel, 1998, Frings et al., 2007, Rothermund et al., 2005).
According to the code occupation account, these partial overlap costs occur because one feature can only be in one event file at a time, meaning that in order to bind for example the distractor feature "LEISE" to the response "red", it might be necessary to "unbind" it first from the response "green".
Another explanation for these costs do not base on the "unbinding" of the prime event file, but on the activation of the feature overlapping between the event files ("LEISE") with onset of the probe target ("red") and the resulting "co-activation" of the bound feature in the prime event file ("green"), which is conflicting with the now required response.
This study investigates whether one feature can actually only be part of one event file, as suggested by the code occupation account. To do so, we adopted the color categorization task (e.g., Rothermund, Wentura, De Houwer, 2005). Participants will position the index, middle and ring fingers of one hand on the keys J, K, and L. Each key will be assigned a color. Participants are asked to respond to the font color of presented words (targest) and ignore the word meanings (distractors). We introduced an intermediate trial (n-1) which occurs between the prime (n-2) and the probe (n). Between n-2 and n, there will be only response alternations, meaning that overlap costs will be examined by repeating / switching the distractor from prime to probe. In half of the trials, trial n-1 will serve the purpose of potentially "destroying" event file n-2. If n-1 partially (with respect to the response feature) overlaps with n-2, then, according to the code occupation account, event file n-2 should be "unbound" and partial overlap costs between n-2 and n disappear. If n-1 however, does not overlap with n-2, partial overlap costs in n with respect to n-2 should be observed. The other half of the trials will serve as a comparison, as trial n-1 will not overlap with n-2, which should result in the typical binding effect in n with regards to n-2.
Distractor words will come from a pool of 750 words and will not repeat between runs
- …
