Language
Lexical Decision Task
Also called: LDT, word-nonword task
Measure how quickly people recognize written words, with common and rare words and pronounceable nonwords.
3-5 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
- 1,412
- Evidence grade
- B
- Registered clinical studies
- 22
What the participant sees
One trial, screen by screen, with the timings the test uses.
500 ms
A cross
Until a key press, up to 2,000 ms
A letter string
- ForJ
- Word or not a word
Which of F and J means word is counterbalanced across sessions and saved with each answer.
Example stimuli
Common word
Zipf frequency 4.53
Rare word
Zipf frequency 2.78
Nonword
Made from "random"
Session structure
Practice
8 trials
Common words
20 trials
Rare words
20 trials
Nonwords
40 trials
Example data
One session, scored by the code that scores your study's sessions.
Word RT
609.4ms
Mean response time of correct answers to words.
Quality check: valid80 scored trials
- Word Accuracy
- 92.5%
- Nonword Accuracy
- 97.5%
- Frequency Effect
- 77.3 ms
- Nonword RT
- 671.1 ms
- Common-Word RT
- 567.6 ms
- Rare-Word RT
- 644.9 ms
- Accuracy
- 95%
- Common words
- Rare words
- Nonwords
- Mean
- Wrong answer
Trial file
The first 6 of 88 rows, with the column names of your study's export.
| participant_id | test_slug | test_version | scoring_version | trial_index | trial_type | response | rt | correct | time_elapsed | anticipation | base_word | correct_response | fixation_ms | frequency_band | key_mapping | length | letter_string | lexicality | locale | nonword_key | paradigm_version | phase | random_seed | response_window_ms | run_id | set_number | stimulus_language | task | timed_out | trial_number | word_key | zipf |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| simulated | lexical-decision | 1.1.0 | 1.0.0 | 0 | html-keyboard-response | f | 584 | true | 2084 | false | chisel | f | 500 | word-j | 6 | chosal | nonword | en | f | 1 | practice | 2282081923 | 2000 | example-run | en | lexical_decision_practice | false | 1 | j | |||
| simulated | lexical-decision | 1.1.0 | 1.0.0 | 1 | html-keyboard-response | j | 659 | true | 4043 | false | j | 500 | word-j | 7 | circuit | word | en | f | 1 | practice | 2282081923 | 2000 | example-run | en | lexical_decision_practice | false | 2 | j | 3.87 | |||
| simulated | lexical-decision | 1.1.0 | 1.0.0 | 2 | html-keyboard-response | f | 743 | true | 6086 | false | naughty | f | 500 | word-j | 7 | saughly | nonword | en | f | 1 | practice | 2282081923 | 2000 | example-run | en | lexical_decision_practice | false | 3 | j | |||
| simulated | lexical-decision | 1.1.0 | 1.0.0 | 3 | html-keyboard-response | j | 571 | true | 7957 | false | j | 500 | word-j | 5 | ounce | word | en | f | 1 | practice | 2282081923 | 2000 | example-run | en | lexical_decision_practice | false | 4 | j | 3.54 | |||
| simulated | lexical-decision | 1.1.0 | 1.0.0 | 4 | html-keyboard-response | f | 701 | true | 9958 | false | welfare | f | 500 | word-j | 7 | delfire | nonword | en | f | 1 | practice | 2282081923 | 2000 | example-run | en | lexical_decision_practice | false | 5 | j | |||
| simulated | lexical-decision | 1.1.0 | 1.0.0 | 5 | html-keyboard-response | j | 664 | true | 11922 | false | j | 500 | word-j | 6 | snatch | word | en | f | 1 | practice | 2282081923 | 2000 | example-run | en | lexical_decision_practice | false | 6 | j | 3.63 |
Scores and how they are computed
RT means response time. Mean RTs use correct responses at or after 100 ms.
- Word RTms
- Mean correct RT to words.
- Lower is faster
- Word Accuracy%
- Share of words answered as words.
- Higher is better
- Nonword Accuracy%
- Share of nonwords answered as not words.
- Higher is better
- Frequency Effectms
- Rare-word minus common-word mean correct RT.
- Positive values mean common words were recognized faster
- Nonword RTms
- Mean response time of correct answers to nonwords.
- Lower is faster
- Common-Word RTms
- Mean correct response time to common words.
- Lower is faster
- Rare-Word RTms
- Mean correct response time to rare words.
- Lower is faster
- Accuracy%
- Share of all test trials answered correctly.
- Higher is better
- Cognate Effectms
- With the cognate setting: noncognate minus cognate mean correct response time.
- Positive values mean cognates were recognized faster
- Cognate RTms
- With the cognate setting: mean correct response time to cognates.
- Lower is faster
- Noncognate RTms
- With the cognate setting: mean correct response time to noncognates.
- Lower is faster
- Cognate Accuracy%
- With the cognate setting: share of cognates answered as words.
- Higher is better
- Noncognate Accuracy%
- With the cognate setting: share of noncognates answered as words.
- Higher is better
What it measures
Measures how quickly people recognize written words. A string of letters appears and participants decide whether it is an English word, pressing F or J. Half the strings are common or rare words and half are pronounceable nonwords.
Constructs
- Visual word recognition
- Lexical access
- Processing speed
Research uses
- Reading and language research
- Bilingualism and second-language research
- Aging and psycholinguistics studies
How the paradigm works
The lexical decision task measures how quickly people recognize written words. A string of letters appears in the middle of the screen and participants decide whether it is an English word, pressing one key for a word and the other for a nonword. Which of F and J means word is drawn from the session's random seed and recorded.
More about the paradigm
Each trial starts with a 500 ms cross, and the letters stay until a response or 2,000 ms. The standard 80 trials hold 20 sets drawn at random from 40. Each set is a common word (Zipf frequency 4.5 or higher), a rare but familiar word (Zipf 2 to 3) and two pronounceable nonwords, all of one length (4 to 7 letters). Nonwords change one or two letters of a real word and are checked against English word lists. Trials run in a recorded random order with no more than three words or three nonwords in a row, after 8 practice trials with feedback.
The main score is the mean response time of correct word answers. The frequency effect is the rare-word minus common-word mean correct response time; positive values mean common words were recognized faster. Words were chosen by fixed rules from the Glasgow Norms and OpenSubtitles word frequencies. The words are English, and sessions run in other languages are flagged for review.
A cognate setting, off by default, is for bilingual participants doing the task in English as their second language. The study chooses their first language: German, Spanish, French, Italian, Dutch or Portuguese. Each set then holds an English cognate of that language, a word spelled the same or nearly so as its usual translation (piano and piano), and a noncognate of the same length and frequency whose translation shares almost no letters (window and ventana), with the set's two nonwords. Each language has 40 pairs, chosen with the orthographic similarity measure of Dijkstra and colleagues (2010): at least 0.75 for cognates and at most 0.25 for noncognates. The cognate facilitation effect is the noncognate minus cognate mean correct response time; positive values mean the cognates were recognized faster, as Dijkstra, Grainger and van Heuven (1999) found. With the setting, the session may run in English or in the chosen first language.
Use and evidence
1,412 papers from 2021 to 2025 name this task in their title or abstract, down 10% compared with 2016 to 2020. That ranks it 18 of 711 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 383 |
| 2024 | 277 |
| 2023 | 246 |
| 2022 | 245 |
| 2021 | 261 |
| 2020 | 276 |
| 2019 | 327 |
| 2018 | 332 |
| 2017 | 303 |
| 2016 | 331 |
| 2015 | 304 |
| 2014 | 320 |
| 2013 | 287 |
| 2012 | 347 |
| 2011 | 314 |
| 2010 | 262 |
| 2009 | 281 |
| 2008 | 268 |
| 2007 | 222 |
| 2006 | 251 |
| 2005 | 182 |
| 2004 | 212 |
| 2003 | 187 |
| 2002 | 179 |
| 2001 | 152 |
| 2000 | 140 |
Use grade A
1,000 or more papers from 2021 to 2025.
Evidence grade B
At least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers.
- Papers on test-retest reliability
- 15
- Papers on reliability or validity
- 251
- Registered clinical studies using it as an outcome
- 22
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 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Trial
- A 500 ms cross, then a letter string until a response or 2,000 ms.
- Keys
- F and J; which key means word is counterbalanced across sessions and saved.
- Items
- 80 trials standard (40 to 160), in sets of a common word, a rare word and two nonwords of one length.
- Words
- Common words Zipf 4.5 or higher and rare words Zipf 2 to 3, chosen by fixed rules from openly licensed norms.
- Nonwords
- Pronounceable strings made by changing one or two letters of real words, checked against English word lists.
- Order
- Random order from the session's seed, with no more than three words or three nonwords in a row.
- Practice
- 8 trials with feedback.
- Cognate words
- A setting, off by default, for bilingual participants doing the task in English as their second language. The study chooses their first language: German, Spanish, French, Italian, Dutch or Portuguese. Each set then holds an English cognate of that language (piano), a noncognate of the same length and frequency (window) and the two nonwords: 40 pairs for each language, with orthographic similarity to the translation of at least 0.75 for cognates and at most 0.25 for noncognates. Scores the cognate facilitation effect, noncognate minus cognate mean correct response time.
The words are English. Sessions run in other languages are flagged for review; with the cognate setting, a session in the chosen first language is not.
Data you get
Every response is saved as it happens. Download scores for all plans, and every trial on yearly plans or with participant credits.
Trial columns for this test
- phase
- practice or test. Scores use test rows.
- trial_number
- Order of the trial within its phase.
- random_seed
- Seed of the session's trial order, to rebuild it.
- letter_string
- The string shown.
- lexicality
- word or nonword.
- frequency_band
- high (common) or low (rare) for words.
- zipf
- Word frequency on the Zipf scale.
- base_word
- The real word a nonword was made from.
- word_key
- The key that meant word in this session, f or j.
- correct_response
- The right key for the trial.
- response
- The key pressed, empty when none.
- rt
- Response time in ms from the moment the screen to answer appeared.
- correct
- Whether the key pressed was the right one.
- timed_out
- No key before the time limit.
- anticipation
- A key under 100 ms, too fast to be a response.
Columns in every export
- participant_id
- Short participant code, the same in every file of the study.
- external_id
- Your own participant ID, when you add one.
- test_version
- Task version the session ran.
- scoring_version
- Scoring version the scores came from.
- customization_status
- standard, or scored_modified when the study changed a setting. A setting that only changes how a test looks, such as a game's Look setting (the game_look trial column), keeps it standard.
- configuration_json
- The task settings the session ran with.
- look
- plain, or game when the study used the game look. A game-look session's percentile is compared with plain-look norms, and the two looks have not yet been shown to agree.
- look_version
- The version of the game layer the session ran. Empty for a plain session.
- is_valid
- The session's quality status (scores file).
- quality_primary_reason
- The main reason when a session needs review or is invalid (scores file).
- started_at
- When the session started (scores file).
- completed_at
- When the session ended (scores file).
Quality checks
Each session is marked valid, needs review or invalid, with the reason.
Scored rows: lexical_decision_test from the scored run, F or J for word or nonword (chance 50%, half the strings are words) within 2,000 ms, correctness from the key for the string. 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, or with the cognate setting other than English and the chosen first language, 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.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Test trials | 80 | 40 to 160 |
| Cognate words | Off (standard) | Off (standard), German and English, Spanish and English, French and English, Italian and English, Dutch and English, Portuguese and English |
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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Lexical Decision Task in a web browser (ConductCognition task version 1.1.0, scoring version 1.0.0; Meyer & Schvaneveldt, 1971), with the standard settings (test trials: 80; cognate words: Off). Word RT was computed as the mean response time of correct answers to words. Mean response times used correct responses at or after 100 ms.
ConductCognition versions: task version 1.1.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Meyer, D. E., & Schvaneveldt, R. W. (1971). Facilitation in recognizing pairs of words: Evidence of a dependence between retrieval operations. Journal of Experimental Psychology, 90(2), 227-234. https://doi.org/10.1037/h0031564
- Balota, D. A., Yap, M. J., Hutchison, K. A., Cortese, M. J., Kessler, B., Loftis, B., Neely, J. H., Nelson, D. L., Simpson, G. B., & Treiman, R. (2007). The English Lexicon Project. Behavior Research Methods, 39(3), 445-459. https://doi.org/10.3758/BF03193014
- Dijkstra, T., Grainger, J., & van Heuven, W. J. B. (1999). Recognition of cognates and interlingual homographs: The neglected role of phonology. Journal of Memory and Language, 41(4), 496-518. https://doi.org/10.1006/jmla.1999.2654The cognate setting: faster recognition of cognates by bilinguals reading their second language.
- Dijkstra, T., Miwa, K., Brummelhuis, B., Sappelli, M., & Baayen, H. (2010). How cross-language similarity and task demands affect cognate recognition. Journal of Memory and Language, 62(3), 284-301. https://doi.org/10.1016/j.jml.2009.12.003English lexical decision by bilinguals, and the orthographic similarity measure that sorts cognates from noncognates.
Source papers
- Meyer DE, Schvaneveldt RW (1971). Facilitation in recognizing pairs of words: evidence of a dependence between retrieval operations. Journal of Experimental Psychology, 90(2):227-234.Paradigm source, Adults
- Balota DA, Yap MJ, Hutchison KA, et al. (2007). The English Lexicon Project. Behavior Research Methods, 39(3):445-459.Paradigm source, Adults
Questions
How long does the Lexical Decision Task take?
About 3 to 5 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?
Word RT, Word Accuracy, Nonword Accuracy, Frequency Effect, Nonword RT, Common-Word RT, Rare-Word RT, Accuracy, Cognate Effect, Cognate RT, Noncognate RT, Cognate Accuracy and Noncognate Accuracy. Each session is also marked valid, needs review or invalid, with the reason.
How is word RT computed?
Mean correct RT to words.
How is the frequency effect computed?
Rare-word minus common-word mean correct RT.
How widely is the Lexical Decision Task used in research?
1,412 papers from 2021 to 2025 name it in their title or abstract, down 10% compared with 2016 to 2020. That ranks it 18 of 711 test families in the ConductCognition test census (use grade A). Source: ConductCognition test census 0.58.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Lexical Decision Task have?
In the published research, 15 papers report its test-retest reliability and 22 registered clinical studies use it as an outcome (evidence grade B in the ConductCognition test census).
Which settings can a researcher change?
Test trials (40 to 160) and cognate words (Off (standard), German and English, Spanish and English, French and English, Italian and English, Dutch and English, Portuguese and English). 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.