Language
Semantic Decision Task
Also called: semantic judgment task, semantic categorization task, living/nonliving decision, animacy judgment, size judgment task, concrete/abstract decision, concreteness judgment
Measure how quickly people retrieve word meaning, deciding whether each word names something living, bigger than a shoebox, or concrete, with words written for each of eight languages.
5-7 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. Try it here first, with no account.
- Papers, 2021 to 2025
- 119
- Evidence grade
- C
- Registered clinical studies
- 3
What it measures
Measures how quickly people decide what a word means. One word at a time appears in the participant's language, and participants press F or J to answer one question per part: is it living or nonliving, bigger or smaller than a shoebox, or concrete or abstract? Scores are the accuracy and time of each decision.
Constructs
- Semantic memory
- Word meaning
- Processing speed
Research uses
- Language and semantic memory research
- Imaging and brain stimulation studies that need a word meaning task
- Aging and bilingualism research comparing decisions across languages
How the paradigm works
The Semantic Decision Task measures how quickly people retrieve what a word means. One word at a time appears in the middle of the screen, in the participant's language, and participants answer one question about what it names, pressing F for one answer and J for the other. Each part asks one question: is it living or nonliving, is it bigger or smaller than a shoebox, or is it concrete or abstract? A study runs all three parts, in an order drawn from the session's seed (the standard), or one of them. The key for living, bigger and concrete is drawn from the seed and kept for the whole session.
More about the paradigm
Each part starts with its question, its keys and 4 practice words with feedback, then 24 test words, 12 for each answer, in a recorded random order with no more than three of one answer in a row. Each word follows a 500 ms cross and stays until a key or 3,000 ms. The living and size words are animals and objects, half of each bigger and half smaller than a shoebox, so the question a part does not ask is balanced within each answer, and no word appears twice. The words were written for this task in each of the eight languages and chosen so that the two answers of each part match on mean length and mean word frequency in that language.
The main score is the mean time of correct answers. The task also reports the accuracy and mean correct time of each decision, and the concreteness effect: abstract minus concrete mean correct time, where positive values mean concrete words were decided faster. Every answer is saved with its word, its answer, its length and its frequency for item analyses. Binder and colleagues (2009) reviewed 120 functional imaging studies of semantic processing and found a left-lateralized network of seven regions; Pexman and colleagues (2017) collected concrete or abstract decisions on 10,000 English words from 321 participants, 1,000 words each. Scores support research and do not diagnose.
A setting, off by default, changes the living or nonliving part's words to typical and atypical members of its two categories, written for each language: animals and plants or fungi for living, still objects and natural things that move, such as clouds and rivers, for nonliving, matched on length and frequency within each answer, with a rule that names no examples. Category decisions are slower for members less typical of their category (Rips, Shoben and Smith 1973; Rosch 1975), and adults are slower and less accurate to call plants living and clouds or rivers nonliving (Goldberg and Thompson-Schill 2009). The setting scores this typicality effect for each answer and both together.
Scores
- Decision Timems
- Mean response time of correct answers, all decisions, answers under 100 ms left out.
- Lower is faster
- Accuracy%
- Share of all test words answered correctly; a word with no key in 3 seconds counts as wrong.
- Higher is better
- Living Decision Timems
- Mean correct response time in the living or nonliving part; empty when the part was not run.
- Lower is faster
- Living Decision Accuracy%
- Share of the living or nonliving part answered correctly.
- Higher is better
- Size Decision Timems
- Mean correct response time in the bigger or smaller than a shoebox part; empty when the part was not run.
- Lower is faster
- Size Decision Accuracy%
- Share of the size part answered correctly.
- Higher is better
- Concreteness Decision Timems
- Mean correct response time in the concrete or abstract part; empty when the part was not run.
- Lower is faster
- Concreteness Decision Accuracy%
- Share of the concrete or abstract part answered correctly.
- Higher is better
- Concreteness Effectms
- Abstract minus concrete mean correct response time in the concrete or abstract part.
- Positive values mean concrete words were decided faster
- Anticipationscount
- Answers under 100 ms.
- Quality check
- Missed Wordscount
- Test words with no key within 3 seconds.
- Quality check
- Test Wordscount
- Test words recorded in the scored run.
- Informational
- Planned Test Wordscount
- Test words the session's plan held: 24 for each part run.
- Informational
- Typicality Effectms
- With the typical and atypical members setting: atypical minus typical mean correct response time in the living or nonliving part. The typical and atypical times and accuracies are also reported.
- Positive values mean atypical members were decided more slowly
- Typicality Accuracy Effect%
- With the typical and atypical members setting: typical minus atypical share answered correctly.
- Positive values mean atypical members were decided less accurately
- Living Typicality Effectms
- With the typical and atypical members setting: plants or fungi minus animals, mean correct response time.
- Positive values mean plants were called living more slowly
- Nonliving Typicality Effectms
- With the typical and atypical members setting: natural things that move minus still objects, mean correct response time.
- Positive values mean moving natural things were called nonliving more slowly
Use and evidence
119 papers from 2021 to 2025 name this task in their title or abstract, up 5% compared with 2016 to 2020. That ranks it 252 of 711 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 38 |
| 2024 | 26 |
| 2023 | 19 |
| 2022 | 15 |
| 2021 | 21 |
| 2020 | 18 |
| 2019 | 30 |
| 2018 | 22 |
| 2017 | 22 |
| 2016 | 21 |
| 2015 | 20 |
| 2014 | 27 |
| 2013 | 19 |
| 2012 | 26 |
| 2011 | 29 |
| 2010 | 19 |
| 2009 | 12 |
| 2008 | 21 |
| 2007 | 17 |
| 2006 | 17 |
| 2005 | 5 |
| 2004 | 10 |
| 2003 | 15 |
| 2002 | 18 |
| 2001 | 10 |
| 2000 | 3 |
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
- 2
- Papers on reliability or validity
- 21
- Registered clinical studies using it as an outcome
- 3
Measured on ConductCognition
Pooled across the studies that feed our platform norms, by device group. A figure needs 50 or more people.
No pooled figure yet.
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 0.58.0, counted 13 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Trial
- A 500 ms cross, then one word until F or J or 3,000 ms.
- Decisions
- Living or nonliving, bigger or smaller than a shoebox, and concrete or abstract, one per part: all three in an order drawn for each session (standard), or one chosen by the study.
- Keys
- F and J. Which key gives living, bigger and concrete is drawn from the session's seed, kept for every part and saved.
- Words
- 24 per part, 12 for each answer, written for this task in each of the eight languages; no word appears twice. The living and size parts use animals and objects, half of each bigger and half smaller than a shoebox.
- Matching
- In each part and language, the words of the two answers match on mean length within 0.34 characters and mean log frequency within 0.10 (OpenSubtitles counts).
- Order
- Random order from the session's seed, with no more than three words with the same answer in a row.
- Practice
- 4 words before each part, 2 for each answer, with feedback for 800 ms.
- Score
- Mean time of correct answers; each decision's time and accuracy; and the concreteness effect, abstract minus concrete time.
- Typical and atypical members
- A setting, off by default, for the living or nonliving part: 24 words in each language, 6 animals and 6 plants or fungi for living, 6 still objects and 6 natural things that move, such as clouds and rivers, for nonliving, matched within each answer on mean length and log frequency, with a rule that names no examples. Scores the typicality effect, atypical minus typical time and typical minus atypical accuracy, for each answer and both together.
Each language has its own words, matched within that language, so decision times compare best within one language. Concrete words name things that can be seen and touched and abstract words name ideas and qualities, classed by that rule rather than by rating norms.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: semantic_decision_test from the scored run, F or J for the two answers of each block's question (chance 50%, half the words take each answer) within 3,000 ms, correctness from the key for 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.
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 words are written for each of the eight languages, so a session in any of them is scored the same way.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Decisions | All three: living, size and concrete (standard) | All three: living, size and concrete (standard), Living or nonliving only, Bigger or smaller than a shoebox only, Concrete or abstract only |
| Typical and atypical members | Off | On or off |
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
- Languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese.
- 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 Semantic Decision Task in a web browser (ConductCognition task version 1.1.0, scoring version 1.1.0; Binder et al., 2009; Pexman et al., 2017), with the standard settings (decisions: All three: living, size and concrete; typical and atypical members: Off). The main score was decision time (mean response time of correct answers, all decisions, answers under 100 ms left out).
ConductCognition versions: task version 1.1.0, scoring version 1.1.0. Every export records the versions each session used, for a methods section.
- Binder, J. R., Desai, R. H., Graves, W. W., & Conant, L. L. (2009). Where is the semantic system? A critical review and meta-analysis of 120 functional neuroimaging studies. Cerebral Cortex, 19(12), 2767-2796. https://doi.org/10.1093/cercor/bhp055The left-lateralized semantic network found across 120 imaging studies of semantic processing.
- Pexman, P. M., Heard, A., Lloyd, E., & Yap, M. J. (2017). The Calgary semantic decision project: Concrete/abstract decision data for 10,000 English words. Behavior Research Methods, 49(2), 407-417. https://doi.org/10.3758/s13428-016-0720-6The concrete or abstract decision, with item-level decision times for 10,000 English words.
- Rosch, E. (1975). Cognitive representations of semantic categories. Journal of Experimental Psychology: General, 104(3), 192-233. https://doi.org/10.1037/0096-3445.104.3.192The typical and atypical members setting: members of a category differ in how typical they are.
- Rips, L. J., Shoben, E. J., & Smith, E. E. (1973). Semantic distance and the verification of semantic relations. Journal of Verbal Learning and Verbal Behavior, 12(1), 1-20. https://doi.org/10.1016/S0022-5371(73)80056-8The typical and atypical members setting: category membership is verified more slowly for members far from their category.
- Goldberg, R. F., & Thompson-Schill, S. L. (2009). Developmental "roots" in mature biological knowledge. Psychological Science, 20(4), 480-487. https://doi.org/10.1111/j.1467-9280.2009.02320.xThe typical and atypical members setting: adults judged life status less accurately and more slowly for plants than animals, and for natural and moving nonliving things than for still artifacts.
Source papers
- Pexman PM, Heard A, Lloyd E, Yap MJ (2017). The Calgary semantic decision project: concrete/abstract decision data for 10,000 English words. Behavior Research Methods, 49(2):407-417.Paradigm source, Adults
- No normative data for this task.Paradigm source, N/A
Questions
How long does the Semantic Decision Task take?
About 5 to 7 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?
Decision Time, Accuracy, Living Decision Time, Living Decision Accuracy, Size Decision Time, Size Decision Accuracy, Concreteness Decision Time, Concreteness Decision Accuracy, Concreteness Effect, Anticipations, Missed Words, Test Words, Planned Test Words, Typicality Effect, Typicality Accuracy Effect, Living Typicality Effect and Nonliving Typicality Effect. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Semantic Decision Task used in research?
119 papers from 2021 to 2025 name it in their title or abstract, up 5% compared with 2016 to 2020. That ranks it 252 of 711 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.58.0, counted 13 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Semantic Decision Task have?
In the published research, 2 papers report its test-retest reliability and 3 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?
Decisions (All three: living, size and concrete (standard), Living or nonliving only, Bigger or smaller than a shoebox only, Concrete or abstract only) and typical and atypical members (On or off). 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, with its words in the same language. 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.