Evidence
Measuring what nobody can report
Ask someone how much effort a screen cost and the answer will be an honest guess, and wrong. Nobody has access to their own attention budget. The seven methods in the first group measure what the body did, without going through opinion. The seven in the second do ask, but they ask in a way that makes the guess comparable, with a calibrated scale, a stopwatch and what the person remembers unprompted.
Two groups, doing different jobs. The first produces the evidence this site cites, and three of the seven you will not operate, because EEG, fMRI and fNIRS ask for a laboratory. The other four fit a product budget, and it is worth knowing the difference before you ask for the money. The second group fits in an ordinary research session, with what you already have. Each method opens on its own page, with what it sees and what it does not.
What is read off the body
7 techniques that read the body before it turns into opinion. Three of them, EEG, fMRI and fNIRS, need a lab; the other four fit a product budget.
7 methods
- EEG and event-related potentials Cognitive effort over time, and, in the ERP cut, the reaction to a single event
- fMRI Which regions activate during the task, with high spatial resolution
- Functional near-infrared spectroscopy The cortex’s oxygen consumption during the task, outside the laboratory
- Eye tracking Where visual attention goes, and in what order
- Pupillometry How much effort the task cost, second by second
- Automated facial expression coding Facial movement during the interaction, and, with heavy caveats, the valence of the reaction
- Galvanic skin response and biosensors Intensity of emotional activation (arousal), not its direction
What you can measure
7 instruments that fit in an ordinary session and need no lab. They are not simplified versions of the laboratory methods. Each one measures a concept the site explains on another page.
7 methods
- NASA Task Load Index Perceived load, separated into six sources
- Task chronometry How long the decision took, and, above all, how that time grows
- Five-second test What remains of a first glance
- Free recall and recognition What was left afterwards, and what evaporated from the middle
- Pictorial emotion scale Valence and arousal, without requiring vocabulary
- Experience sampling The experience at the instant it happens, in the real world
- Extreme-user interview What the average user struggles to verbalise
Before citing any of these
Where evidence runs out
None of these methods authorises any conclusion on its own. Three limits apply to all of them, and ignoring those is the difference between using science and using it as decoration.
Where evidence runs out →