Social Cognition
Approach-Avoidance Task
Also called: approach avoidance task, AAT, manikin task, stimulus-response compatibility task
Measure automatic approach and avoidance: participants walk a small figure toward or away from pleasant and unpleasant words with the arrow keys, scored by how much faster movements are when they match the word's valence.
7-9 minTimed: desktop or laptop with a mouse or trackpad
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 251
- Evidence grade
- C
- Registered clinical studies
- 28
What it measures
Measures the pull to approach pleasant words and avoid unpleasant ones. A small figure stands above or below a word, and participants walk it toward the word or away with the arrow keys. In one part they approach pleasant words and in the other unpleasant words; responses are usually faster when the movement matches the word's valence.
Constructs
- Approach and avoidance tendencies
- Automatic evaluation
- Affective stimulus-response compatibility
Research uses
- Research on automatic approach and avoidance of pleasant and unpleasant words
- Studies of anxiety, eating and substance use that measure approach biases
- Within-study comparison of evaluative and letter-case versions
How the paradigm works
The Approach-Avoidance Task measures how quickly people move toward pleasant things and away from unpleasant ones. A small stick figure appears above or below the middle of the screen, and 750 ms later a word appears in the middle. Participants press the up or down arrow key to walk the figure toward the word or away from it. In one part they approach pleasant words and avoid unpleasant ones; in the other part the rule is reversed. Which part comes first is set at random for each session. Each part starts with 8 practice trials with feedback, followed by 64 test trials.
More about the paradigm
The task grew out of studies by Solarz (1960) and Chen and Bargh (1999), whose participants pulled a lever toward themselves or pushed it away in response to words and were faster when the movement matched the word's valence. De Houwer and colleagues (2001) replaced the arm movement with a figure that key presses move toward or away from the word, so the movement is defined by the word rather than the body, and Krieglmeyer and Deutsch (2010) compared this manikin version with two joystick versions. A meta-analysis found a reliable effect when participants respond to the word's valence and almost none when valence is irrelevant to the task (Phaf, Mohr, Rotteveel and Wicherts, 2014), so the standard version asks for valence. The 32 pleasant and 32 unpleasant words are chosen by rule from the Glasgow Norms and matched pair by pair on length, frequency and arousal. A setting switches to a version in which half the words are in capitals and the rule is about letter case, as in De Houwer and colleagues (2001).
Scores
- Approach Biasms
- Median correct response time on incompatible trials (approaching unpleasant or avoiding pleasant words) minus compatible trials (approaching pleasant or avoiding unpleasant words).
- Positive when movements that match the word's valence are faster
- Median Compatible RTms
- Median correct response time when approaching pleasant or avoiding unpleasant words.
- Informational
- Median Incompatible RTms
- Median correct response time when approaching unpleasant or avoiding pleasant words.
- Informational
- Approach Minus Avoid, Pleasant Wordsms
- Median correct time to approach pleasant words minus the time to avoid them.
- Negative when pleasant words are approached faster
- Approach Minus Avoid, Unpleasant Wordsms
- Median correct time to approach unpleasant words minus the time to avoid them.
- Positive when unpleasant words are avoided faster
- Error Rate, Compatible%
- Share of compatible trials not answered correctly. A missed response counts as an error.
- Informational
- Error Rate, Incompatible%
- Share of incompatible trials not answered correctly. A missed response counts as an error.
- Usually higher than on compatible trials
- Accuracy%
- Proportion of test trials answered with the key the rule asks for.
- Informational
- Compatible Part Firstflag
- 1 when the part that asks for approaching pleasant words ran first, 0 when it ran second. Empty in the letter-case version.
- For analysing the bias by part order
- Approach Bias, Part 1ms
- The approach bias within the first part, in the letter-case version, where each part has both kinds of trial.
- Informational
- Approach Bias, Part 2ms
- The approach bias within the second part, in the letter-case version.
- Informational
- Anticipationscount
- Responses within 100 ms of the word's onset.
- Informational
Use and evidence
251 papers from 2021 to 2025 name this task in their title or abstract, up 6% compared with 2016 to 2020. That ranks it 99 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 67 |
| 2024 | 40 |
| 2023 | 44 |
| 2022 | 46 |
| 2021 | 54 |
| 2020 | 46 |
| 2019 | 48 |
| 2018 | 55 |
| 2017 | 50 |
| 2016 | 38 |
| 2015 | 40 |
| 2014 | 34 |
| 2013 | 21 |
| 2012 | 20 |
| 2011 | 14 |
| 2010 | 8 |
| 2009 | 6 |
| 2008 | 3 |
| 2007 | 1 |
| 2006 | 1 |
| 2005 | 0 |
| 2004 | 0 |
| 2003 | 1 |
| 2002 | 3 |
| 2001 | 0 |
| 2000 | 1 |
Use grade C
100 to 299 papers from 2021 to 2025.
Evidence grade C
At least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies.
- Papers on test-retest reliability
- 6
- Papers on reliability or validity
- 74
- Registered clinical studies using it as an outcome
- 28
How the grades work
Use: A is 1,000 or more papers from 2021 to 2025, B 300 to 999, C 100 to 299, D 30 to 99 and E fewer than 30. Evidence: A means at least 25 papers report its test-retest reliability and at least 25 registered clinical studies use it as an outcome; B, at least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers; C, at least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies; D, little published measurement evidence found. The counts measure use. They include every paper that names the task in its title or abstract. Grades describe the published research on this task, not ConductCognition's version of it.
Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Display
- A word in the middle of the screen and a small stick figure 110 px above or below it, drawn in code.
- Trial
- A blank screen for 500 ms and the figure alone for 750 ms, then the word until a key or 3,000 ms.
- Response
- The up and down arrow keys walk the figure toward the word or away from it, in 6 steps over 300 ms; it then stands for 200 ms. Response time runs from the word's onset to the first key.
- Parts
- Two parts with opposite rules: toward pleasant and away from unpleasant words, and the reverse, in an order set at random for each session.
- Trial mix
- 64 trials per part (standard; 32 to 96 in steps of 16): half pleasant and half unpleasant words, with the figure above the word on half of each. Each test word appears once per part, in the other position in the other part.
- Words
- 32 pleasant and 32 unpleasant English words chosen by rule from the Glasgow Norms, matched pair by pair on length, frequency and arousal.
- Practice
- 8 trials with feedback before each part, with words of their own.
- Options
- A letter-case version: half the words are in capitals and the rule is about letter case, so valence is irrelevant to the task.
The standard version asks participants to respond to valence. In a meta-analysis this gave a reliable effect, and versions in which valence is irrelevant gave almost none (Phaf et al., 2014). The words are English; sessions run in other languages are flagged for review.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: approach_avoidance_test from the scored run, both parts, the up or down arrow key that walks the figure toward or away from the word as the part's rule asks (chance 50%, each key correct on half of the trials) within 3,000 ms of the word. Below chance invalidates and not above chance is flagged. One key on 90% of answers invalidates when accuracy is not above chance. Missed responses and responses under 100 ms follow the timed-test rule. A session run in a language other than English is flagged for review.
Why these rules
Accuracy significantly below chance (one-sided exact test, p < .05) cannot come from guessing; it means the response mapping was reversed or the task was not followed (Slick, Sherman and Iverson 1999 treat below-chance forced-choice performance as definite response bias). Chance-level accuracy can be a real result in impaired participants, so it is only flagged. One option on at least 90% of answered trials, when the correct options are balanced and accuracy is not above chance, means the answers did not depend on the items (the task counterpart of long-string responding in surveys; Curran 2016). Missed responses and responses under 100 ms keep the platform's timed-test rule: more than max(3, 25%) missed responses or more than max(2, 20%) responses under 100 ms invalidate, and any is flagged. A response under 100 ms cannot be a reaction to the stimulus. The word lists are English. A session run in another language shows English words to a participant who may not read English fluently, so it is flagged for review.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Response rule | The word's meaning: toward pleasant or unpleasant words (standard) | The word's meaning: toward pleasant or unpleasant words (standard), Letter case: toward words in capitals or in small letters |
| Trials per part | 64 (standard) | 32, 48, 64 (standard), 80, 96 |
Yearly plans can change settings within these limits. Free and pay-as-you-go studies use the standard settings.
Need a setting this test does not have?
Devices and languages
- Timed test: participants need a desktop or laptop with a mouse or trackpad. Participant devices
- Instructions and screens: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. Words and sentences: English.
- Available on: Every plan, including the free plan.
- No percentile. Scores and quality flags only. How norms work
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Approach-Avoidance Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Solarz AK, 1960; Chen M, Bargh JA, 1999; De Houwer J, Crombez G, Baeyens F, Hermans D, 2001), with the standard settings (response rule: The word's meaning: toward pleasant or unpleasant words; trials per part: 64). The main score was approach bias (median correct response time on incompatible trials (approaching unpleasant or avoiding pleasant words) minus compatible trials (approaching pleasant or avoiding unpleasant words)).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Solarz, A. K. (1960). Latency of instrumental responses as a function of compatibility with the meaning of eliciting verbal signs. Journal of Experimental Psychology, 59(4), 239-245. https://doi.org/10.1037/h0047274
- Chen, M., & Bargh, J. A. (1999). Consequences of automatic evaluation: Immediate behavioral predispositions to approach or avoid the stimulus. Personality and Social Psychology Bulletin, 25(2), 215-224. https://doi.org/10.1177/0146167299025002007
- De Houwer, J., Crombez, G., Baeyens, F., & Hermans, D. (2001). On the generality of the affective Simon effect. Cognition and Emotion, 15(2), 189-206. https://doi.org/10.1080/02699930125883The manikin task and the letter-case version.
- Krieglmeyer, R., & Deutsch, R. (2010). Comparing measures of approach-avoidance behaviour: The manikin task vs. two versions of the joystick task. Cognition and Emotion, 24(5), 810-828. https://doi.org/10.1080/02699930903047298
- Rinck, M., & Becker, E. S. (2007). Approach and avoidance in fear of spiders. Journal of Behavior Therapy and Experimental Psychiatry, 38(2), 105-120. https://doi.org/10.1016/j.jbtep.2006.10.001
- Phaf, R. H., Mohr, S. E., Rotteveel, M., & Wicherts, J. M. (2014). Approach, avoidance, and affect: A meta-analysis of approach-avoidance tendencies in manual reaction time tasks. Frontiers in Psychology, 5, 378. https://doi.org/10.3389/fpsyg.2014.00378Effect sizes by instruction.
Source papers
- De Houwer et al. (2001): the manikin task. Krieglmeyer and Deutsch (2010): comparison with joystick versions. Phaf et al. (2014): a meta-analysis of approach-avoidance tasks.Paradigm source, Adults
Questions
How long does the Approach-Avoidance Task take?
About 7 to 9 minutes, as listed in the test library.
What device does a participant need?
Timed test: participants need a desktop or laptop with a mouse or trackpad. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Approach Bias, Median Compatible RT, Median Incompatible RT, Approach Minus Avoid, Pleasant Words, Approach Minus Avoid, Unpleasant Words, Error Rate, Compatible, Error Rate, Incompatible, Accuracy, Compatible Part First, Approach Bias, Part 1, Approach Bias, Part 2 and Anticipations. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Approach-Avoidance Task used in research?
251 papers from 2021 to 2025 name it in their title or abstract, up 6% compared with 2016 to 2020. That ranks it 99 of 395 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Approach-Avoidance Task have?
In the published research, 6 papers report its test-retest reliability and 28 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Response rule (The word's meaning: toward pleasant or unpleasant words (standard), Letter case: toward words in capitals or in small letters) and trials per part (32, 48, 64 (standard), 80, 96). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
The instructions and screens are in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. The words and sentences stay in English, so sessions in other languages are flagged for review. English is on every plan, and the other languages come with yearly plans. Have a native speaker read the machine-translated text before launch.
Is this a clinical test?
No. ConductCognition is for research use only. It runs, scores and exports the test; clinical interpretation is outside the platform.
Every test: the full test list.
Run this test in your next study
Start free with 10 free participants on any test, then use participant credits or a yearly plan.