Perception Lab
Our Perception Lab is a browser-based perceptual task tool for students, teaching demos, and introductory research. Run classic perception tasks, compare conditions, and export usable data without cobbling together visual stimuli by hand and then acting surprised when the whole thing goes strange.
It includes four core tasks in one clean interface: Müller-Lyer Judgement, Ebbinghaus Size Judgement, Ponzo Length Judgement, and Visual Search. Use Free Play to explore the tasks quickly, or switch to Project Mode for participant instructions, practice trials, condition labels, and export-ready results.
Personal use is free. Commercial and institutional use requires a licence which covers all our Labs.
Task Setup
Configure the task for this participant. Keep it tidy. That usually pays off later.
Task Instructions
Show these instructions before the task begins.
A short 3-trial practice round will begin first. These practice trials are not included in the final results.
Task Complete
Participant: | Condition:
What Perception Lab Suite Does
Our Perception Lab gives you four classic perceptual tasks in one clean browser-based tool. It is designed for students, teaching demos, and introductory research where you need something more structured than improvised slides and more manageable than heavy lab software. Use Free Play to explore the tasks quickly, or switch to Project Mode for participant instructions, practice trials, condition labels, and export-ready data.
The Four Tasks
Müller-Lyer Judgement measures line-length judgements under misleading visual context. Participants decide which line appears longer, making it useful for demonstrating how surrounding cues can distort apparently simple perceptual decisions.
Ebbinghaus Size Judgement measures size perception under contextual influence. Participants judge which centre circle looks larger while surrounding circles quietly interfere, making it useful for illustrating contrast effects and contextual bias in visual judgement.
Ponzo Length Judgement measures perceived length under depth and perspective cues. Participants decide which line looks longer when converging context lines suggest distance, making it useful for demonstrations of size constancy, spatial interpretation, and perceptual bias.
Visual Search measures target detection across feature and conjunction displays. Participants decide whether a target is present or absent, making it useful for simple attention work and for comparing the relative ease of different search conditions.
How to Use It for a Simple Study
Choose one task and keep the design manageable. Select an independent variable such as time pressure, screen size, distraction, practice, or display condition. Keep the testing conditions as consistent as possible, run each participant through the same task, then export the results and compare control accuracy, bias rate, reaction time, or target detection performance depending on the task you used.
How to Interpret Your Results
Perception Lab gives you trial-by-trial CSV data, which is much more useful than one neat-looking final score that tells you almost nothing. The sections below explain what the download contains, what to look for first, and how to read the patterns without immediately declaring that perception itself has collapsed.
Your CSV includes one row per trial. That usually means participant ID, task name, condition label, trial number, stimulus type, response, correct response, accuracy, reaction time in milliseconds, and timestamp.
In plain terms, each row tells you what the participant saw, how they responded, whether the response was correct, and how long it took.
Open the CSV in Excel or Google Sheets. The first sensible step is to sort or filter by condition label so you can compare groups or testing conditions without turning the file into a wall of numbers and regret.
After that, focus on the outcome columns that matter for the task you used, especially accuracy, response choice, and reaction time.
Accuracy: Start by checking how often participants responded correctly. If one condition lowers accuracy reliably, that is usually the clearest first sign that the perceptual manipulation mattered.
Reaction time: Then look at how quickly participants responded on correct trials. Slower responses can suggest that the stimulus was harder to interpret, even if accuracy stayed fairly stable.
Error patterns: If your task involves different response categories, check whether certain mistakes happen more often than others. Sometimes the interesting result is not just that participants got things wrong, but that they got them wrong in a consistent direction.
Condition-specific difficulty: Look for whether one type of stimulus, contrast level, ambiguity level, or perceptual manipulation reliably produces worse performance than the others.
Keep the design simple. If you ran a high-contrast versus low-contrast condition, a distraction versus no-distraction condition, or an easy versus ambiguous stimulus comparison, sort the sheet by condition label and compare the main outcome columns directly.
First compare accuracy: if one condition produces more errors, that is often the strongest effect.
Then compare reaction time: if accuracy is similar but one condition is slower, the perceptual demand may still be higher.
Then check the error pattern itself: if participants confuse one category with another in a repeated way, that is usually more interesting than a few scattered mistakes.
A consistent pattern across many trials is worth taking seriously. One bizarre response in the middle of an otherwise tidy dataset is usually just one bizarre response.
A lower accuracy score or a slower reaction time does not automatically prove some grand theory about human perception. It may reflect stimulus difficulty, distraction, fatigue, hesitation, or participants simply not paying close attention.
This is a browser-based teaching and introductory research tool. It is very useful for classroom work, demos, and undergraduate projects, but it is not a tightly controlled lab environment with specialist presentation hardware.
Open the CSV, sort by condition, then compare accuracy first and reaction time second. After that, look for whether the same kinds of mistakes keep appearing under the same perceptual conditions.
In most cases, that will tell you far more than staring at one overall average and trying to persuade yourself it contains wisdom.
Example Study Ideas
This tool works well for straightforward introductory designs, including:
screen size or viewing distance on Müller-Lyer bias
brief practice effects on Ebbinghaus size judgements
time pressure on Ponzo control accuracy
feature versus conjunction search performance in Visual Search
quiet versus distracted testing conditions across any of the four tasks
A Quick Note on Precision
Perception Lab is a browser-based tool designed for teaching, demos, exploratory work, and introductory student research. It is useful, clean, and a great deal less irritating than building perceptual tasks from scratch, but it is not intended to replace specialist experimental software in high-precision lab settings.
Well suited to undergraduate work. For PhD-level precision, that is usually where the bigger software budgets and the slightly haunted lab computers begin.
