Task Background Info
In the Task-Switching (TS) task, participants were presented with a series of one of four possible stimuli and asked to respond to the stimulus based on the task cue presented prior to, and above, the image. The four stimuli included a red triangle, red circle, green triangle, and green circle. Participants switched between responding to the image’s color (i.e., red or green) or the shape (i.e., triangle or circle). Cues presented included either “SHAPE” or “S” on trials where participants were expected to respond to the shape of the stimulus; cues presented included either “COLOR” or “C” on trials where participants were expected to respond to the color of the stimulus. On 33% of trials the instructions switched, such that participants were instructed to switch from responding from shape to color, or vice versa. On 67% of trials, the instructions remained the same but the cues changes. This task is designed to measure the changes in reaction time between trials requiring vs. not requiring a switch in responding.
Participants completed a total of 96 trials, for a total run time of 6 min 52 s.
For the practice run, participants completed 10 trials of the task. Before reading task instructions, experimenters would mention to participants that this task is difficulty for people, and that a lot of time people need to redo the demo more than once in order to get comfortable with the task, in order to reduce frustration on behalf of the participants. As the button order was counterbalanced across participants, instructions below represent one such button order (and were switched for other participants).
This is a demo of the task in which you will learn to press the LEFT (blue) button or the RIGHT (yellow) button in response to color and shapes. The colors will be red or green. The shapes will be circles or triangles. Each trial starts with a cue indicating whether you should respond to color or shape. Try to respond as fast and as accurately as you can for each trial.
If the cue is “COLOR” or “C” respond to the color. Press the LEFT button if the color is RED. Press the RIGHT button if the color is GREEN. If the cue is “SHAPE” or “S” respond to the shape. Press the LEFT button if the shape is a TRIANGLE. Press the RIGHT button if the shape is a CIRCLE.
So, again, if the cue is “COLOR” or “C”: Press the LEFT button if the color is RED. Press the RIGHT button if the color is GREEN. If the cue is “SHAPE” or “S”: Press the LEFT button if the shape is a TRIANGLE. Press the RIGHT button if the shape is a CIRCLE.
Scan Instructions This task is the one where you have to respond to the color or the shape of images we show you on the screen as fast and as accurately as possible. At the top, you’ll see a cue that tells you what to respond to. If the cue says “COLOR” or “C,” you should respond to the color, and if it says “SHAPE” or “S” you should respond to the shape. When it says to respond to color, press the first button if it’s red, and the second button if it’s green. When it says to respond to shape, push the first button if it’s a triangle, and the second button if it’s a circle. Again, that’s red for button 1 and green for button 2, and triangle for button 1 and circle for button 2. Any questions?
Participants saw on the screen When cued with “S” or “Shape,” respond to the shape of the image. FIRST button = Circle SECOND button = Triangle
When cued with “C” or “Color,” respond to the color of the image. FIRST button = Red SECOND button = Green
Respond as quickly and as accurately as possible.
Scoring Behavioral Data AND Creating Onset Files
There is just one TS scoring script, which makes onset files and behavioral output at same time.
1. Scripts can either be run so that you do all subjects of a given group at once, or do one subject at a time. There are comments in the code that instruct you to indicate first with group (e.g., CONTROLS) you will be doing. By default, it will run all subjects in the selected group unless you choose to comment in a couple lines and enter in a particular subject's ID (e.g., CNP_10150).
2. in matlab, run 'TASKSWITCH_onsets_analyze.m'
For BEHAVIORAL DATA, this will either create behavioral output (in SUBJ/behav/TASKSWITCH) for each subject in your group or just the subject you specified. This script creates 2 files:
a. It adds to the 'GROUP_TASKSWITCH_group_output.txt' file, by adding a line for each subject with their summary scores.
b. It creates a 'SUBJ_TASKSWITCH_behav_output.txt' file in the subject's behav/TASKSWITCH directory (as well as a trial-by-trial summary file).
The file created under 2a will be used for analysis of behavioral data collected during the TS scan. In most cases, this data will be provided after query.
For ONSET FILES:
a. The following onset files will be created in SUBJ/behav/TASKSWITCH, either for each subject in your group or just the subject you specified.
Each onset file is in the same 3-column format as other tasks, with Onset time, Duration, and Weight--with Duration = RT on that trial, and Weight fixed at 1 for all trials.
b. The script also creates a 'GROUP_TASKSWITCH_log.txt' file that flags any subject with an empty onset file for a given trial type. In this case, the subject would need to be flagged.
Running First Level Analyses
1. There are two scripts for the two TASKSWITCH models, although model2 is favored now.
There are separate scripts for each GROUP (e.g., 'run_level1_TASKSWITCH_CONTROLS.sh').
Each of these scripts creates an individualized .fsf file for each subject (which is then stored in /space/raid2/data/poldrack/CNP/scripts/designs/TASKSWITCH), then runs it, and the output is stored in the subject's analysis/TASKSWITCH directory.
2. To run each script, open it in emacs.
a. As is commented into the script, you need to comment in certain sections depending on whether the subject was scanned at BMC vs. CCN. This is in two places, at the top and bottom of the script.
b. Enter in the subject(s) ID(s) on a new line of code (e.g., 'for id in 11131, do').
c. Save and submit to grid to run.
List of Models
level1_TASKSWITCH_model2 (this is the preferred model now)
Model description and contrasts
-- INTERNAL CHECKS
1. Task total CongShort trials
2. Task total CongLong trials
3. Task total InconShort trials
4. Task total InconLong trials
5. Task total NSshort trials
6. Task total Sshort trials
7. Task total NSlong trials
8. Task total Slong trials
9. button set
10. Task total correct
11. Task Percent Correct responses
12. Task mean RT
13. Task SD RT
-- The following 5 variables (A-E)
A. Total number of trials
B. Total Correct
C. Total Percent Correct responses
D. Mean RT
E. SD RT
For each of the following Categories of trial types:
CSI Delay-Short (34-38)
CSI Delay-Long (39-43)
Congruent Switch Short Delay (44-48)
Congruent NoSwitch Short Delay (49-53)
Congruent Switch Long Delay (54-58)
Congruent NoSwitch Long Delay (59-63)
Incongruent Switch Short Delay (64-68)
Incongruent NoSwitch Short Delay (69-73)
Incongruent Switch Long Delay (74-78)
Incongruent NoSwitch Long Delay (79-83)
Link back to LA5C page.