Behavioral Tasks & Analyses Programsused for different studies |
||
| These tasks are scripts, which can be run using a compiled TCL script environment called paradigms2004.exe, which can be downloaded here and a self-extracting zip here. | ||
|
Paradigms Behavioral Tasks GCC Behavioral Tasks Programs to analyze data sets from the behavioral tasks:
|
Rock Paper Scissors Task: This task is used to determine whether subjects are able to select responses that are associated with wins and avoid responses that are associated with losses; whether subjects are able to change their response selection when the associated outcome contingencies change. This task is based on the well-known Rock Paper Scissors game. For the task, the typical rules apply: paper beats rock, rock beats scissors, and scissors beat paper. The subjects were instructed that they were playing against the computer and were told to maximize their total point account (1 point for a win, 0 points for a tie, and -1 point for a loss). The probability of reinforcement, i.e. beating the computer (e.g. subject chooses paper, computer selects rock, subject gains one point), was pre-determined for each response within a trial block (see Figure 2). Unbeknownst to the subject and without changing trial duration or inter-trial interval, the preferred, even, and worst response were switched every 20 trials. For example, when the subject selects scissors (preferred response) during the first block the computer will select paper in 9/10 encounters. When the subject selects rock (even response), the computer will select rock in 8/10 encounters, paper in 1/10 encounters and scissors in 1/10 encounters. When the subject selects paper (worst response), the computer will select scissors in 9/10 encounters. A total of 120 trials are presented; therefore the “best response” block switches 6 times after 20 consecutive trials in the following order: scissors, paper, rock, scissors, rock, paper. The main dependent variable was the proportion of response that is associated with wins 9/10 times during each trial block. Download Analysis Software here.
Visual Working Memory Task Visual Working Memory Task: Purpose: Central to human information processing is the concept of a limited-capacity memory system for the short-term retention and manipulation of information. Baddeley proposed a multi-component model of working memory in which memory for visual and verbal information is stored separately in different subsystems and are controlled by a central executive. The most notable characteristic of WM is that its storage capacity is highly limited. Results from functional neuroimaging studies of verbal WM are consistent with its conceptualization as a phonological loop. In comparison, visual WM depends on the number objects but not on the number of different object features. Using different experimental manipulations it was estimated that, on average, subject can hold 3-4 visual objects in mind. Description: Based on a memory task previously described by Phillips & Singer,Vogel and Luck developed a paradigm that allows one to easily estimate visual working memory capacity and to parametrically manipulate working memory load in terms of number of items to be recalled. In addition, the executive component, i.e. the degree to which the temporarily stored information has to be manipulated, is minimal in this paradigm. In this task colored dots of various set sizes (sample) are presented simultaneously at random locations on the computer screen for a brief period of time. After a short delay, another set of colored dots (test) is presented simultaneously at the same random locations. Both sets are either the same or the sets differ on one feature, e.g. color, in one of the objects. In this sample-test dot array visual working memory task, the subject has to indicate whether the sample and the set are the same or different and, therefore, has to encode both object and spatial information for correct responding. All stimulus arrays are presented within a 10 cm x 10 cm region on a video monitor with a gray background. The color of each object is randomly selected from a set of 7 colors. When a feature is changed between sample and test, the new value is selected at random from all of the possible feature values. Each trial consists of a 100 msec sample followed by a 900 msec delay period and then a 2000 msec test array. The sample and test arrays are identical except that one feature of one item in the test array is different from that of the corresponding item in the sample array on 50% of the trials. The subject responds by pressing one of two buttons on each trial to indicate whether the sample and test arrays were the same or different.
Download Analysis Software here.
Go/Nogo Task: Participants view a serial
presentation of blue shapes on a computer screen, comprised of large
circles, small circles, large squares, and small squares.
Each stimulus appears for 200 msec.
Subjects are instructed to press a button each time a large
circle, small circle, or large square (go stimulus) appears, and not to
press the button when a small square (nogo stimulus) is shown.
The intertrial interval is 1500 msec. The task is presented in 18 blocks of 10 trials each. There are two block types (9 of each): one consisting of salient cue (SC) sequences and one of nonsalient cue (NC) sequences. Subjects are not informed that there are different types of sequences. In both types, when a large square appears, it is frequently followed by a small square (79% of the time across all SC sequences, 76% across all NC sequences), thus serving not only as a go stimulus but also as a cue that the nogo stimulus is likely to appear next. In the SC sequences, large squares appear relatively frequently (about 27% of stimuli), so there are many pairings of this cue with the nogo stimulus. In the NC sequences, large squares appear less frequently (15% of stimuli), resulting in fewer pairings. Thus, while a given appearance of the cue has about the same probability of being followed by a nogo stimulus in the two sequence types, in the SC sequences cues appear more often in the first place, leading to more cue-nogo pairings and correspondingly more salient cue-nogo associations.
Download Analysis Software here.
Risky Gains Task Risky Gains Task: Subjects are presented with a sequence of three numbers in ascending order (20, 40, and 80). Each number is displayed onscreen for one second and, if the subject presses a button while that number is displayed, he/she receives that number of points along with immediate positive visual and auditory feedback. When a 40 or 80 appears, however, there is a chance that it will appear in an alternate color, along with immediate negative feedback signaling a loss of 40 or 80 points, respectively. When this occurs, the trial ends immediately (i.e. the subject may not make a response). Each trial lasts 3.5 seconds regardless of the subject’s selection or punishment. Download Analysis Software here.
Two Choice Prediction Task Two Choice Prediction Task: In this task, a house is presented on the computer screen, which is flanked by a person to the left and right. The goal for the subject is to decide on which side of the house a car will be presented. Each trial is self-paced to maximize self-determined action, thus the subject determines the number of trials by the latency to select a response. Immediately following the subject’s response, the car is presented for 300 msec on the far left or right side of the screen, which provides the feedback whether the prediction was correct, i.e. the person and car are on the same side, or incorrect, i.e. the person and car are on opposite sides. Immediately, after the presentation of the car, the next trial begins. Unbeknownst to the participants, the computer determined the response based on the participant’s selection. Three error rate block types were presented: a high (20% of responses were “correct”), a chance-level (50% of responses were “correct”), and a low (80% of responses were “correct”) error-rate block type. Download Analysis Software here.
Gamble Task
CERT Task
CERT Task:
There are several complex issues associated with the estimation
of decision weights (Kuehberger, 1998) . Depending on the manner of inquiry, individuals can exhibit
inconsistent preferences, show strong framing effects, and overvalue
either reward magnitude or reward probability.
Tversky and Kahneman suggested a procedure akin to psychophysical
experiments of sound or signal threshold detection (Tversky & Kahneman,
1992)
. As shown in the Figure
above, this procedure arrives at the best estimate for the certainty
equivalent by iteratively narrowing the interval that is likely to
contain the certainty equivalent value. Subjects were instructed prior to the MRI about this task in the following manner: "In many situations you have to decide between a sure thing or a gamble, something that may or may not pay out. In this task you select between a chance event and a sure thing. You may see: 1: 1 in 10 chance of winning $100 or 2: A sure win of $10. If you prefer option 1, please press the LEFT key. If you prefer option 2, please press the RIGHT arrow key. After your selection the options disappear and a new trial begins.” Subjects are presented with a choice between a gamble (option 1) or a sure thing (option 2). Seven different prospects are presented with a probability, p, and a value, x, i.e. (p,x) = (0.002, 10000), (0.06, 2000), (0.1,500), (0.55,700), (0.7, 150), (0.9, 400), (0.99, 5000). For each prospect, the subject completes eight trials to successively narrow the range of the certainty equivalent (CE) value. As shown in Figure 1, each trial began with the presentation of the gamble and the sure thing. The subjects have no time limit to respond to the presentation. Once the subject made a response an “x” appeared next to the selected options for 2 seconds, which was followed by a blank screen for 4 seconds before the next trial begins.
Temporal Discounting Task: Subjects are informed that they have to make preference judgments about hypothetical rewards and are instructed as follows: “You will be asked to decide between a gain NOW and a gain sometime in the FUTURE. The future gain may be several days from now or as remote as 10 years from now. For example, you may see: (1) A gain of $512 in one year; (2) A gain of $10 now. If you prefer option 1, please press the LEFT key. If you prefer option 2, please press the RIGHT key.” The task consists of 6 blocks of 8 preference judgment trials each. On each trial, subjects are presented with a choice between a future reward (option 1) and an immediate reward (option 2). Option 1 is the same for all trials of a given block, with a block-specific delay in days, d, and reward magnitude in dollars, x, i.e. (d, x) = (5, 506), (30, 476), (180, 524), (365, 512), (1095, 520), (3650, 488) for blocks 1-6, respectively. The reward magnitude of option 1 is varied slightly over blocks within the same order of magnitude so as not to alter the rate of discounting but to prevent stereotyped responding, i.e. always selecting the current or delayed option. Across blocks the mean greater reward is $504.3 (range: 476 to 524). Option 2, the immediate reward, varied in magnitude from trial to trial within each block according to a rule described below. The pairing of a particular delay with a particular reward is the same for every subject. Subject responses enable us to successively narrow the range of the DE value. This procedure is analogous to the psychophysical determination of a stimulus detection threshold and consists of systematically varying the magnitude of the immediate reward to find which amount delivered immediately would be preferred equally to the delayed reward associated with each block.
Both the CERT task and the Temporal Discounting Task have been revised so as to work with consecutive bi-section instead of the original triple section. This has the advantage of needing fewer trials but is not auto-corrective, i.e. if the subjects makes a decision to go with a lower interval, there is no turning back.
Download the new version of CERT and DISC here.
Emotional Face Processing Task – block design: This is a slightly modified version of the emotion face assessment task (see (Hariri et al 2002c) ). During each 5 second trial, a subject is presented with a target face (on the top of the computer screen) and two probe faces (on the bottom of the screen) and is instructed to match the probe with the same emotional expression to the target by pressing the left or right key on a button box (see above). A block consists of six consecutive trials where the target face is either angry, fearful, or happy (Ekman et al 1983) . During the sensorimotor control task subjects are presented with 5-second trials of ovals or circles in an analogous configuration and instructed to match the shape of the probe to the target. Each block of faces and of the sensorimotor control task is presented three times in a pseudo-randomized order. A fixation cross is interspersed between each block. For each trial, response accuracy and response latency are obtained. The regressors of interest (see analysis pathway below) mark the different block types (angry, fear, happy, and sensorimotor control). This paradigm, written in Presentation, can be downloaded here.
A non-presentation version can be downloaded here. Programs to analyze behavioral data: These programs are written in gcc and are usually used from the command line or can be run by a batch program (niteowl.exe). The input for these programs are the output of the generated data files from the different tasks. The output is a tab delimited text file, which can be read into a spreadsheet or statistics program.
|
