Attention has EEG correlates. It does not have an EEG score.

"Focus" is not one thing. Psychology distinguishes sustained attention, selective attention, divided attention and alerting, and they behave differently and are measured by different tasks. Any single number labelled focus has already collapsed that distinction before it starts.

What the research actually supports

Alpha and spatial attention. The most solid finding in the area. When you attend to one side of space, alpha over the opposite hemisphere decreases while alpha over the same side increases — the brain appears to suppress the region processing the ignored side. This is robust, replicated, and measured across many trials in controlled conditions rather than read off a single moment.

Alpha and eyes. Alpha drops on eye opening and on visual engagement. Real, and much cruder than it sounds — it tracks visual input more than mental effort.

Frontal theta and cognitive control. Theta over the midline frontal region increases during tasks demanding conflict monitoring and working memory. Genuine and reasonably reproducible, and typically measured with electrodes at the top of the head rather than the forehead.

Event-related potentials. The P300, a positive deflection around 300 ms after a meaningful stimulus, varies with attention reliably enough to be a research workhorse. It requires many repetitions of a controlled stimulus and averaging across them — it does not exist as a continuous readout.

Where the popular claims overreach

Theta/beta ratio. Widely promoted as an attention index and studied heavily in ADHD. It did not hold up as a diagnostic marker: effect sizes shrank with better-controlled studies, and it is not accepted as a diagnostic test.

Beta as concentration. Beta rises with alertness and also with muscle tension, and the forehead sits on a muscle. A "focus" figure derived from forehead beta will reliably increase when you frown, which is not the same as concentrating.

A single-moment score. The effects above are group averages across many trials. Individual variation is large enough that a single reading from one person at one moment carries little information — the honest version of most of this research is a difference of a few per cent, visible only in aggregate.

Neurofeedback, briefly and fairly

Neurofeedback trains people to alter their own brain activity by showing it to them in real time. The literature is large and the quality is uneven. Reviews of ADHD neurofeedback consistently find that effects reported by parents and teachers who know which group a child is in are considerably larger than effects measured under proper blinding — and blinded assessments often show little or nothing.

This does not mean it never works. It means the evidence is weaker than the enthusiasm, that placebo and expectation account for a meaningful share of reported benefit, and that any claim of clinical benefit deserves the standard of evidence any other clinical claim would need.

What is more honestly measurable

Not focus itself, but several things that shape it and can be measured properly:

Drowsiness. Slow rolling eye movements, longer eye closures and increased blink duration are clear, large signals — an EOG story more than an EEG one, and forehead electrodes see them well.

Time of day. Alertness follows a circadian rhythm with a well-documented afternoon dip, and when your dip falls is a personal fact worth knowing. That is a separate guide.

The room. Carbon dioxide and noise have measurable effects on cognitive performance, they vary enormously through a day, and unlike a brain score they are straightforward to measure and straightforward to fix. See CO₂ and concentration and how noise affects concentration.

There is an argument buried in that list. If the goal is to concentrate better, the environmental measurements are both more trustworthy and more actionable than the brain ones — you can open a window.