Approach Avoidance Task - AAT - Joystick

Technical Manual

Script Author: Katja Borchert, Ph.D. (katjab@millisecond.com), Millisecond

Created: January 17, 2012

Last Modified: January 13, 2025 by K. Borchert (katjab@millisecond.com), Millisecond

Script Copyright © Millisecond Software, LLC

Background

This script implements the Approach-Avoidance Task (AAT), a test of implicit behavioral avoidance tendencies. This script uses a joystick for response making and is thus limited to Windows.

References

Wiers, R.W., Rinck, M., Dictus, M. & Wildenberg van den, E. (2009). Relatively strong automatic appetitive action-tendencies in male carriers OPRM1 G-allele. Genes, Brain and Behavior, 8, 101-106.

Duration

6 minutes

Description

Participants are asked to do a categorization task of portrait vs. landscape pictures (of 4 different categories, e.g. alcohol, soda etc). If the picture is a landscape picture, they are asked to push the joystick away from body which shrinks the picture (=AVOID). If the picture is a portrait picture, they are asked to pull the joystick towards body which enlarges the picture (=APPROACH). Pull (Approach) and Push (Avoid) responses in regard to landscape and portrait formats are counterbalanced across participants by groupnumber.

Procedure

format/direction counterbalanced by group number (odd vs even) (Inquisit Lab: manual input of subject id and group id at beginning of task)
1. runs a trial sequence generation block via helper script AATSequenceGenerator.iqjs
2. runs a practice block
3. runs the AAT test block

PRACTICE:
• PracticeStimuli: gray rectangle in landscape and portrait format
• 10 trials (5 landscape and 5 portrait formats in random order)
• gives error feedback

TEST BLOCK:
• 80 test trials (4 categories x 2 formats x 10 repetitions) in a semi randomized order
(see below for more info on generating the trial sequence)
• intertrial interval of 300ms

TRIAL SEQUENCE:
Wiers et al (2009) used a semi randomized order of all 80 test stimuli with the following two constraints:
1. no more than three consecutive stimuli of the same category
2. no more than three consecutive stimuli of the same format

a helper script "AATSequenceGenerator" generates a trialsequence that fulfills those constraints.

1. The time to generate this sequence DOES differ across participants
Safeguards are in place to prevent the script to take longer than 5s (default setting).
If the script cannot produce a sequence that fullfills all constraints within that time
limit or within the pre-established number of attempts (default: 1000) - whichever comes first- ,
a pre-generated sequence will be randomly selected from the pool of pre-generated sequences.
In the majority of cases, however, a new sequence should be found within ms.

2. The code is specific to running 8 categories
If more categories should be run, go to section 'BLOCKS' and follow instructions to use a complete randomized
algorithm (without constraints).

Stimuli

Wiers et al (2009): categorization of 4 types of stimuli
1) alcohol pictures
2) color and shape matched appetitive stimuli (e.g pictures of soda bottles)
3) good pictures containing people and animals
4) bad pictures containing people and animals

• 10 stimuli per category (each of these in landscape and portrait format)
• The stimuli used in this script are based on the original ones used.
Millisecond thanks Dr. Wiers for generously sharing his material!

Just a word of caution: the stimuli are culturally specific and may not work for your study.
To substitute your own stimuli, go to
script aat_stimuli.iqjs and replace the images

Instructions

The instructions are not original to Wiers et al (2009). They are provided by Millisecond
as htm/html pages and can be edited by changing the provided htm/html files.
To edit htm/html-files: open the respective documents in simple Text Editors such as TextEdit (Mac)
or Notepad (Windows).
See script aat_joystick_instructions_inc.iqjs

Summary Data

File Name: aat_joystick_summary*.iqdat

Data Fields

NameDescription
inquisit.version Inquisit version number
computer.platform Device platform: win | mac |ios | android
startDate Date the session was run
startTime Time the session was run
subjectId Participant ID
groupId Group number
sessionId Session number
elapsedTime Session duration in ms
completed 0 = Test was not completed
1 = Test was completed
expCondition Used for counterbalancing experimental conditions
1: push for landscape, pull for portrait
2: pull for landscape, push for portrait
propCorrectOverall Proportion correct across all test trials
meanRTOverall Mean latency (in ms) of correct responses across all test trials
sdRT Standard deviation of correct responses across all test trials
medianRTOverall Median latency (in ms) of correct responses across all test trials
aatDiffScoreCat1 Corrected AAT Difference Scores for Cat1/Cat2 (here: alcohol images)
Diff Scores = median latency of correct trials for Push trials - median latency of correct trials for Pull trials
latency only for initial response
interpretation
Positive Scores = Pushing slower than pulling => "APPROACH"
Negative Scores = Pulling slower than pushing => "AVOID"
aatDiffScoreCat2 Corrected AAT Difference Scores for Cat1/Cat2 (here: non-alcohol images)
aatDiffScoreCat3 Corrected AAT Difference Scores for Cat1/Cat2 (here: other positive images)
aatDiffScoreCat4 Corrected AAT Difference Scores for Cat1/Cat2 (here: other negative images)
sequence The experimental trial sequence generated for the participant

Raw Data

File Name: aat_joystick_raw*.iqdat

Data Fields

NameDescription
build Inquisit version number
computer.platform Device platform: win | mac |ios | android
date Date the session was run
time Time the session was run
subject Participant ID
group Group number
session Session number
expCondition Used for counterbalancing experimental conditions
1: push for landscape, pull for portrait
2: pull for landscape, push for portrait
blockcode The name the current block (built-in Inquisit variable)
blocknum The number of the current block (built-in Inquisit variable)
trialcode The name of the currently recorded trial (built-in Inquisit variable)
trialnum The number of the currently recorded trial (built-in Inquisit variable)
trialnum is a built-in Inquisit variable; it counts all trials run
even those that do not store data to the data file.
currentTrial The name of the currently running Trial Sequence (e.g. AAT1)
stimulus The presented stimulus image
targetCategory The targetcategory of the targetstimulus, 1 - 4
1) alcohol pictures
2) color and shape matched appetitive stimuli (e.g pictures of soda bottles)
3) positive pictures containing people and animals
4) negative pictures containing people and animals
targetFormat The targetFormat of the targetstimulus (l = landscape; p = portrait)
correctResp Stores the correct response for the current stimulus
initialResponse Stores the original response to the stimulus (PUSH vs. PUL)
accuracy Stores the correctness of the initial response (1 = correct; 0 = error)
finalResponse Stores the final response: PUSH vs. PULL
changeDirection Stores how often participant changed direction during zooming (0 = no change)
rt Stores the latency of the initial response (in ms)
completeRT Stores how long it takes until the joystick is fully extended in whatever direction (in ms)

Parameters

The procedure can be adjusted by setting the following parameters.

NameDescriptionDefault
interTrialInterval Sets the intertrialinterval for the AAT trials in ms 300