Processing Speed
Mirror Reading Task
Also called: mirror-reversed reading, mirror reading skill, reading mirror-reversed text, transformed text reading, reversed text reading
Measure perceptual skill learning: participants read mirror-reversed word triads aloud over five blocks, scored by the speed-up on new and repeated triads. No sound is recorded.
8-14 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
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how quickly people learn to read mirror-reversed words. Participants read groups of three mirrored words aloud and press the space bar when done, over several blocks. The score is the speed-up from the first block to the last. No sound is recorded, so the test cannot hear the words.
Constructs
- Perceptual skill learning
- Procedural learning
- Repetition priming
- Reading
Research uses
- Skill learning and memory systems research
- Comparison of general skill and item-specific learning
- Within-study comparison of groups and conditions
How the paradigm works
The Mirror Reading Task measures how quickly people learn to read mirror-reversed words, a perceptual skill that amnesic patients learned at the rate of matched controls and kept for at least three months (Cohen and Squire, 1980). Words that come back again and again are read faster still, which separates the general skill from what is learned about particular words (Deweer and colleagues, 1993; Ofen-Noy, Dudai and Karni, 2003).
More about the paradigm
Each screen shows a triad of three words in mirror writing: the whole line is reflected, so the first word is on the right and every letter faces the other way. The participant reads the words aloud from right to left and presses SPACEBAR once all three are read. A block holds 10 triads: the same 5 repeated triads in every block and 5 new ones, in an order drawn from the session's seed, as in Panouillères and colleagues (2016). The standard session has 5 blocks after one practice triad, all in one sitting, where the classic studies trained over days. The words are common nouns written in each language, with no b, d, p or q in Latin letters.
The test records no sound: the microphone is not used, so it cannot hear the words, and misreadings are not scored. A triad under 150 ms per word is too fast to have been read aloud and is left out. The primary score is the speed-up on new triads, how much less time the median new triad took in the last block than in the first; the repeated triads' speed-up, their advantage over new triads in the last block and a power-law learning slope are also given. A speed-up depends on how slowly a person starts, so it is read across a group.
Scores
- New Triad Speed-Up%
- How much less time the median new triad took in the last block than in the first, in percent.
- Higher means more learning of the skill; read across a group
- Repeated Triad Speed-Up%
- How much less time the median repeated triad took in the last block than in the first, in percent.
- Higher means more learning; read across a group
- Repetition Benefitms
- Median new triad minus median repeated triad in the last block.
- Positive values mean the repeated words were read faster
- New Triad Learning Slopescore
- Slope of the log median time of new triads on the log block number, a power law.
- More negative is faster learning
- First Block New Triad Timems
- Median time of the new triads in the first block.
- Lower is faster
- Last Block New Triad Timems
- Median time of the new triads in the last block.
- Lower is faster
- First Block Repeated Triad Timems
- Median time of the repeated triads in the first block.
- Lower is faster
- Last Block Repeated Triad Timems
- Median time of the repeated triads in the last block.
- Lower is faster
- Median Triad Timems
- Median time of every timed triad.
- Lower is faster
- Triads Too Fast To Readcount
- Triads under 150 ms per word, too fast to have been read aloud, left out of the times.
- More than max(2, 20%) of the triads invalidates; any is flagged
- Anticipationscount
- Triads ended under 100 ms after they appeared.
- Expected to be 0
- Triads Timedcount
- Triads that count in the times.
- Informational
- Planned Triadscount
- Triads in the session's blocks.
- Informational
- Blockscount
- Blocks in the session.
- Informational
Use and evidence
1 papers from 2021 to 2025 name this task in their title or abstract. That ranks it 490 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 0 |
| 2024 | 0 |
| 2023 | 0 |
| 2022 | 1 |
| 2021 | 0 |
| 2020 | 2 |
| 2019 | 0 |
| 2018 | 0 |
| 2017 | 0 |
| 2016 | 2 |
| 2015 | 0 |
| 2014 | 4 |
| 2013 | 2 |
| 2012 | 0 |
| 2011 | 3 |
| 2010 | 1 |
| 2009 | 2 |
| 2008 | 3 |
| 2007 | 0 |
| 2006 | 4 |
| 2005 | 1 |
| 2004 | 1 |
| 2003 | 1 |
| 2002 | 1 |
| 2001 | 1 |
| 2000 | 1 |
Use grade E
Fewer than 30 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
- 0
Note on the counts: counted with specific phrases, since 'mirror reading' is also a method in biblical studies; undercounts.
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 0.22.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Triads
- Three common nouns on one line, the whole line reflected so the first word is on the right and every letter faces the other way, 34 px on a light panel. The words are written in each of the eight languages, with no b, d, p or q in Latin letters.
- Reading
- The participant reads the three words aloud from right to left and presses SPACEBAR once all three are read; the triads come 1,000 ms apart. A triad's time runs from its display to the press.
- Blocks
- 5 blocks of 10 triads (standard; 3 to 5 as a setting): the same 5 repeated triads in every block and 5 new ones, in an order drawn from the session's seed with no more than three of a kind in a row. The seed also picks which of the 30 triads repeat.
- Practice
- 1 practice triad, then its time and its words in normal print.
- Recording
- None. The microphone is not used and no sound is recorded, so the test cannot hear the words, and misreadings are not scored.
- Score
- The speed-up on new triads: how much less time the median new triad took in the last block than in the first, in percent. Also the speed-up on repeated triads, the repeated triads' advantage in the last block and a power-law learning slope. A triad under 150 ms per word is too fast to have been read aloud and is left out.
The test hears nothing: it times the presses, so a word misread or skipped is not caught, and a misreading or a pause is part of the time. A speed-up depends on how slowly a person starts and rests on five new triads a block, so it is read across a group. The classic studies trained over several days; this version measures the speed-up within one session. Triads too fast to have been read aloud are flagged, and more than max(2, 20%) of them invalidate. The words were chosen for this test in each language, so compare people who read in the same language. 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: mirror_reading_triad, one per triad of the scored run's blocks (practice excluded), timed from the triad's display to SPACEBAR. A reading time has no chance level, and the test hears nothing, so misread words are not counted. A triad under 150 ms per word is too fast to have been read aloud: more than max(2, 20%) of the triads that fast invalidate, and any is flagged. A run short of its planned triads is flagged. A hidden tab withholds the percentile.
Why these rules
Saying a word takes longer than 150 ms even at the fastest rate of speech, and mirror-reversed words take far longer, so a triad that fast was ended without reading it; the platform's timed-test proportion, more than max(2, 20%), invalidates. The test records no sound, so it cannot tell words read aloud from words read silently or skipped. Reading times run while the tab is hidden, so a hidden tab withholds the percentile.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Blocks | 5 | 3 to 5 |
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 Mirror Reading Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Cohen & Squire, 1980; Deweer, Pillon, Michon & Dubois, 1993; Panouillères, Tofaris, Brown & Jenkinson, 2016), with the standard settings (blocks: 5). The main score was new triad speed-up (how much less time the median new triad took in the last block than in the first, in percent).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Cohen, N. J., & Squire, L. R. (1980). Preserved learning and retention of pattern-analyzing skill in amnesia: Dissociation of knowing how and knowing that. Science, 210(4466), 207-210. https://doi.org/10.1126/science.7414331Mirror reading as a skill learned at a normal rate in amnesia and kept for at least three months.
- Deweer, B., Pillon, B., Michon, A., & Dubois, B. (1993). Mirror reading in Alzheimer's disease: Normal skill learning and acquisition of item-specific information. Journal of Clinical and Experimental Neuropsychology, 15(5), 789-804. https://doi.org/10.1080/01688639308402596Unique and repeated word triads, separating the skill from learning of particular words.
- Ofen-Noy, N., Dudai, Y., & Karni, A. (2003). Skill learning in mirror reading: How repetition determines acquisition. Cognitive Brain Research, 17(2), 507-521. https://doi.org/10.1016/S0926-6410(03)00166-6Repeated words learned faster than new ones, with power functions fitted to reading times over blocks.
- Panouillères, M. T. N., Tofaris, G. K., Brown, P., & Jenkinson, N. (2016). Intact acquisition and short-term retention of non-motor procedural learning in Parkinson's disease. PLOS ONE, 11(2), e0149224. https://doi.org/10.1371/journal.pone.0149224One triad per screen read aloud from right to left, blocks of 5 repeated and 5 new triads, and words with no b, d, p or q.
- Poldrack, R. A., & Gabrieli, J. D. E. (2001). Characterizing the neural mechanisms of skill learning and repetition priming: Evidence from mirror reading. Brain, 124(1), 67-82. https://doi.org/10.1093/brain/124.1.67
Source papers
- Panouillères MTN, Tofaris GK, Brown P, Jenkinson N (2016). Intact acquisition and short-term retention of non-motor procedural learning in Parkinson's disease. PLOS ONE, 11(2):e0149224.Paradigm source, Adults
Questions
How long does the Mirror Reading Task take?
About 8 to 14 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?
New Triad Speed-Up, Repeated Triad Speed-Up, Repetition Benefit, New Triad Learning Slope, First Block New Triad Time, Last Block New Triad Time, First Block Repeated Triad Time, Last Block Repeated Triad Time, Median Triad Time, Triads Too Fast To Read, Anticipations, Triads Timed, Planned Triads and Blocks. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Mirror Reading Task used in research?
1 papers from 2021 to 2025 name it in their title or abstract. That ranks it 490 of 495 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.22.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Mirror Reading Task have?
In the published research, 0 papers report its test-retest reliability and 0 registered clinical studies use it as an outcome (evidence grade D in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Blocks (3 to 5). 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.