What it does
This page runs three classic attention and memory experiments in your browser, times every answer and summarises the results the way a psychology lab report needs them. It is meant for class demonstrations, student practicals and trying out a design before building it in lab software. Nothing is uploaded or stored: the trials live in the page until you close or reload it, and you can download them as a CSV first.
- Stroop: a colour word (RED, GREEN, BLUE, YELLOW) appears in coloured ink and you name the ink. When the word and ink disagree (incongruent), most people are slower and make more errors.
- Flanker: five arrows appear and you answer the direction of the middle one. Neighbours pointing the other way slow you down.
- N-back: letters appear one at a time and you press Space when a letter is the same as the one 1, 2 or 3 places back.
It works well with the reliability and agreement, the sample size and power calculator and the publication plot maker.
How to use it
- Pick a task and set the number of trials. For Stroop and flanker, choose the share of congruent trials (50 % is usual; a low share makes the effect bigger). For n-back, choose N, the share of targets and the pace.
- Press Start. On a computer, answer with the keys shown (R, G, B, Y for Stroop; the left and right arrow keys or F and J for flanker; Space for n-back). On a phone or tablet, tap the buttons under the screen. Esc or Stop ends a run early and keeps the trials done so far.
- Each Stroop and flanker trial shows a + for half a second, then the stimulus until you answer or 2 seconds pass. An n-back letter shows for half a second and you can answer until the next one.
- Read the summary, change the trimming rules if you like, and download the trials as CSV or the summary as an image, PDF or Word file.
What the numbers mean
- Reaction time (RT) is measured from the frame in which the stimulus was drawn to the key press or touch, using the browser’s high-resolution clock. It is approximate: screens, keyboards and touch screens add a delay of a few to a few tens of milliseconds. That delay is about the same on every trial, so differences between conditions are much more reliable than single times.
- Mean and median RT use correct trials only. Answers faster than the fastest RT kept (150 ms by default) are treated as anticipations, and slower than the slowest kept (3000 ms) as lapses. The optional SD rule also drops RTs more than 2, 2.5 or 3 standard deviations from their condition’s mean.
- The congruency effect is the mean RT on incongruent trials minus congruent trials. The table also gives medians, SDs and accuracy for each condition.
- For n-back, d′ compares hits (Space on a target) with false alarms (Space on a non-target), with 0.5 added to each count so a perfect score still gives a finite value. The first N letters cannot be targets and are left out.
Worked example
Every run uses a fresh random order, so your numbers will differ. In a typical 40-trial Stroop run with 50 % congruent trials, a student might average about 650 ms on congruent and 720 ms on incongruent trials, a Stroop effect of about 70 ms, with 95 % correct. The summary shows “Stroop effect = 70 ms”, the table lists both conditions, and the chart plots each trial’s RT so slow outliers stand out.
Tips
- Practise with 8 trials first, then do a real run of 40 to 100.
- Use the same device for everyone you compare, and close other tabs.
- The Stroop task needs colour vision; use the flanker task for anyone who finds the colours hard to tell apart.
Frequently asked questions
Is it accurate enough for research? For teaching and pilot work, yes. Published studies usually use dedicated software with checked display timing.
Where are my results kept? Only in this page. Reloading clears them, so download the CSV first.
Was the maths checked? Yes. The summary statistics were compared with pandas and the n-back d′ with SciPy.
Where the numbers come from
- Task designs: Stroop (1935), Eriksen and Eriksen (1974) flanker task, Kirchner (1958) n-back. Trial lists are generated at random in the page.
- Summary formulas: standard descriptive statistics; n-back d′ with the log-linear correction (Hautus 1995).
Standard scientific values, such as constants, atomic weights and melting points, are facts found in every textbook; we name where ours were checked. If you are an author or publisher and want a value, a credit or a table changed or removed, tell us and we will act quickly.
Last updated 2026-11-09.