Executive Function
Acquired Equivalence Task
Also called: acquired equivalence paradigm, equivalence learning task, associative learning and transfer task
Measure generalization from learned equivalence: participants learn by feedback which pets four people have, then meet pairings never trained, scored as the generalization error ratio.
5-10 minUntimed
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
- 33
- Evidence grade
- D
- Registered clinical studies
- 2
What it measures
Measures generalization from learned equivalence. Participants learn by feedback which pets four people have, where two people share a pet, then meet pairings never trained. The main score is the share of errors on those new pairings (the generalization error ratio), with errors on the learned pairings and during learning next.
Constructs
- Associative learning
- Generalization
- Acquired equivalence
- Learning from feedback
Research uses
- Research on hippocampal and basal ganglia contributions to learning
- Parkinson's disease, schizophrenia and aging research
- Within-study comparison of learning and generalization
How the paradigm works
The acquired equivalence task measures whether people generalize what they learn about one stimulus to another that has been paired with the same outcome (Myers, Shohamy, Gluck and colleagues, 2003). Each trial shows a person and two pets, and the participant chooses the pet the person has, with feedback. Learning runs in three stages: first person A1 has pet X1 and person B1 has pet Y1; then A2 also has X1 and B2 also has Y1, so A1 and A2 become equivalent, as do B1 and B2; then A1 also has X2 and B1 also has Y2. A test without feedback then shows the learned pairings and two never trained: A2 with X2 or Y2, and B2 with X2 or Y2. Choosing X2 for A2 and Y2 for B2 follows from the learned equivalence. Myers and colleagues found that people with Parkinson's disease learned more slowly but generalized well, while older adults with hippocampal atrophy learned well but generalized poorly.
More about the paradigm
Each stage repeats its pairings in blocks, every pairing once with each pet on each side, in a seeded order, and ends after 8 correct answers in a row (12 in the third stage) or after 8 blocks (4 to 8 in the settings). Nothing on screen marks a new stage. The pets stay as buttons until a choice by click, tap, or the F and J keys, with no time limit, and Correct or Incorrect shows for 1 second. The test has 3 blocks (2 to 4) of the 8 pairings, each pet on each side: 48 trials, 12 of them new. The four faces are drawn by code for this task: each has one of two values on hair, head shape and shirt, and any two share exactly one, as the source's faces did. The pets are four Noto emoji animals. Which face and pet play which role comes from the session's seed. A practice of 4 trials with another person comes first. About 5 to 10 minutes.
The main score is the generalization error ratio: the share of test trials with a new pairing answered against the equivalence. The retrieval error ratio on the learned pairings comes next, then errors and trials during learning, the stages that reached their run of correct answers, and median test choice times. A session whose learning did not reach every stage's run, or that made more than 25% errors on the learned pairings in the test, is flagged for review. Scores support research and do not diagnose.
Scores
- Generalization Error Ratio%
- Share of test trials with a never-trained pairing (A2 or B2 with X2 and Y2) answered against the learned equivalence.
- Lower means more generalization; 0.5 is chance
- Retrieval Error Ratio%
- Share of test trials with a learned pairing answered wrongly.
- Lower is better; above 0.25 is flagged
- Learning Error Ratio%
- Share of wrong choices over the three learning stages.
- Lower is better
- Learning Errorscount
- Wrong choices over the three learning stages.
- Lower is better
- Learning Trialstrials
- Trials the three learning stages took; at least 28.
- Fewer means faster learning
- Stages Learnedcount
- Learning stages that reached their run of correct answers (8, 8 and 12).
- 3 means every stage was learned
- Learning Complete0/1
- 1 when all three learning stages reached their run of correct answers.
- 0 is flagged for review
- Median Retrieval Timems
- Median choice time on learned pairings in the test.
- Informational
- Median Generalization Timems
- Median choice time on new pairings in the test.
- Informational
- Anticipationscount
- Presses within 100 ms of the pets appearing, in learning and the test. They are not taken as choices.
- Quality check
- Test Choices Madecount
- Test choices in the scored run.
- Fewer than planned is flagged for review
- Planned Test Trialscount
- Test trials in the session: 16 per test block.
- Informational
Use and evidence
33 papers from 2021 to 2025 name this task in their title or abstract, up 27% compared with 2016 to 2020. That ranks it 286 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 9 |
| 2024 | 6 |
| 2023 | 9 |
| 2022 | 5 |
| 2021 | 4 |
| 2020 | 5 |
| 2019 | 6 |
| 2018 | 4 |
| 2017 | 5 |
| 2016 | 6 |
| 2015 | 1 |
| 2014 | 3 |
| 2013 | 6 |
| 2012 | 5 |
| 2011 | 4 |
| 2010 | 4 |
| 2009 | 8 |
| 2008 | 7 |
| 2007 | 5 |
| 2006 | 1 |
| 2005 | 3 |
| 2004 | 1 |
| 2003 | 3 |
| 2002 | 5 |
| 2001 | 3 |
| 2000 | 1 |
Use grade D
30 to 99 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 1
- Registered clinical studies using it as an outcome
- 2
Note on the counts: includes animal studies.
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 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- People and pets
- 4 faces drawn by code for this task, each with one of two values on hair, head shape and shirt, any two sharing exactly one, and 4 Noto emoji pets. Which face and pet play which role is drawn from the session's seed.
- Learning
- Three stages: A1 has X1 and B1 has Y1; then A2 has X1 and B2 has Y1 as well; then A1 has X2 and B1 has Y2 as well. Each stage keeps the earlier pairings.
- Blocks
- Each block shows every pairing of the stage once with each pet on each side, in a seeded order. A stage ends after 8 correct answers in a row (12 in the third stage), or after 8 blocks (standard; 4 to 8).
- Trial
- The pets stay until a choice by click, tap, or the F and J keys, with no time limit. In learning, Correct or Incorrect shows for 1,000 ms, then a 500 ms blank. Presses under 100 ms are not taken.
- Test
- 3 blocks (standard; 2 to 4) of the 6 learned pairings and the 2 new ones, A2 and B2 with X2 and Y2, each pet on each side: 48 trials, 12 of them new, with no feedback.
- Practice
- 4 trials with another person and two other pets, one always correct.
The source used drawings of people and colored fish; this test uses faces and pets of its own, so its error ratios are not directly comparable with published ones. The scores support research and do not diagnose.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: acquired_equivalence_test from the scored run, a choice between two pets with no time limit and no feedback. The chance checks use the learned pairings, which have a trained answer (chance 50%); the new pairings have none and are the measure. Accuracy below chance invalidates and accuracy not above chance is flagged. One side on 90% of choices invalidates when accuracy is not above chance; the trained pet is on each side equally often. Answers under 200 ms on at least half the choices invalidate. A learning stage that ended without its run of correct answers, more than 25% errors on the learned pairings in the test, and a run short of its planned choices are flagged for review and withhold the percentile.
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). An answer under 200 ms is faster than a simple visual reaction, so it cannot reflect the item. Following response-time effort screening for untimed tests (Wise and Kong 2005), the session is invalid only when at least half the answers are that fast. Generalization to new pairings means something only once the learned pairings are known: Myers and colleagues (2003) judged transfer only in participants who reached the learning criterion, and left out one who made many errors on the learned pairings in the test.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Most blocks per learning stage | 8 | 4 to 8 |
| Test blocks | 3 | 2 to 4 |
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
- Untimed test: runs on desktops, laptops, tablets and phones. Participant devices
- Languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese.
- Available on: Every plan, including the free plan.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Acquired Equivalence Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Myers, Shohamy, Gluck et al., 2003; Collie et al., 2002), with the standard settings (most blocks per learning stage: 8; test blocks: 3). The main score was generalization error ratio (share of test trials with a never-trained pairing (A2 or B2 with X2 and Y2) answered against the learned equivalence).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Myers, C. E., Shohamy, D., Gluck, M. A., Grossman, S., Kluger, A., Ferris, S., Golomb, J., Schnirman, G., & Schwartz, R. (2003). Dissociating hippocampal versus basal ganglia contributions to learning and transfer. Journal of Cognitive Neuroscience, 15(2), 185-193. https://doi.org/10.1162/089892903321208123The task: three learning stages and their criteria, the test of new pairings, and slower learning in Parkinson's disease against poorer transfer with hippocampal atrophy.
- Collie, A., Myers, C. E., Schnirman, G. M., Wood, S. J., & Maruff, P. (2002). Selectively impaired associative learning in older people with cognitive decline. Journal of Cognitive Neuroscience, 14(3), 484-492. https://doi.org/10.1162/089892902317361994The task's earlier form, in older people with cognitive decline.
- Bódi, N., Csibri, É., Myers, C. E., Gluck, M. A., & Kéri, S. (2009). Associative learning, acquired equivalence, and flexible generalization of knowledge in mild Alzheimer disease. Cognitive and Behavioral Neurology, 22(2), 89-94. https://doi.org/10.1097/WNN.0b013e318192ccf0The task in mild Alzheimer disease.
- Braunitzer, G., Őze, A., Eördegh, G., Pihokker, A., Rózsa, P., Kasik, L., Kéri, S., & Nagy, A. (2017). The development of acquired equivalence from childhood to adulthood: A cross-sectional study of 265 subjects. PLoS ONE, 12(6), e0179525. https://doi.org/10.1371/journal.pone.0179525Learning, retrieval and generalization error ratios from age 3 to adulthood.
Source papers
- Myers CE, Shohamy D, Gluck MA, Grossman S, Kluger A, Ferris S, Golomb J, Schnirman G, Schwartz R (2003). Dissociating hippocampal versus basal ganglia contributions to learning and transfer. Journal of Cognitive Neuroscience, 15(2):185-193.N 48, Older adults
- Braunitzer G, Őze A, Eördegh G, Pihokker A, Rózsa P, Kasik L, Kéri S, Nagy A (2017). The development of acquired equivalence from childhood to adulthood: A cross-sectional study of 265 subjects. PLoS ONE, 12(6):e0179525.N 265, 3 to 52
- No normative data for this task.Paradigm source, N/A
Questions
How long does the Acquired Equivalence Task take?
About 5 to 10 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops, tablets and phones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Generalization Error Ratio, Retrieval Error Ratio, Learning Error Ratio, Learning Errors, Learning Trials, Stages Learned, Learning Complete, Median Retrieval Time, Median Generalization Time, Anticipations, Test Choices Made and Planned Test Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Acquired Equivalence Task used in research?
33 papers from 2021 to 2025 name it in their title or abstract, up 27% compared with 2016 to 2020. That ranks it 286 of 395 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.14.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Acquired Equivalence Task have?
In the published research, 0 papers report its test-retest reliability and 2 registered clinical studies use it as an outcome (evidence grade D in the ConductCognition test census).
Which settings can a researcher change?
Most blocks per learning stage (4 to 8) and test blocks (2 to 4). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
Participants can take this test in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. 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.