How does the Dual N-Back Test work?
Every 3 seconds a square lights up on a 3 × 3 grid for half a second and your browser speaks a letter. Press Position when the square is in the same place as N steps back, and Sound when the letter is the same as N steps back. Some steps match in one stream, some in both, and most in neither. You choose N: 1-back, 2-back or 3-back.
Each run has 20 scored steps (the first N can't match yet), with 6 matches in each stream, 2 of them on the same steps. The letters are H, J, L, O, Q, R, S and Y, picked so that no two rhyme when spoken. They come from your device's own speech voice, so they sound different on different devices; the square shows for 500 ms whatever the voice. A tick or a cross on each button confirms every press.
How is the Dual N-Back Test scored?
Your score is the share of right decisions across both streams. Every scored step is one decision for Position and one for Sound: press when it matched, hold back when it didn't. That is 40 decisions in a run. Beside it are the matches you caught in each stream and your false alarms, presses on steps that didn't match. Because most steps don't match, holding back earns a lot on its own:
| How you play | Score |
|---|---|
| Never press either button | 70% |
| Press both buttons on every step | 30% |
| Perfect on positions, never press Sound | 85% |
| Perfect on both | 100% |
What is the difference between n-back and dual n-back?
N-back tracks one stream; dual n-back tracks two at once, here where a square appears and which letter you hear. Each stream alone is the N-Back Test's rule; holding both, through your eyes and your ears, is what makes it hard. The n-back task goes back to Wayne Kirchner (1958), who used it to study age differences in remembering information that keeps changing.
Does dual n-back training raise IQ?
Probably not. Susanne Jaeggi, Martin Buschkuehl, John Jonides and Walter Perrig (2008) reported that training on a demanding working memory task raised scores on fluid intelligence tests, more with more training, which made dual n-back famous. Thomas Redick and colleagues (2013) gave young adults 20 sessions of adaptive dual n-back, against an active placebo group, and found no gain on any of the ability tests. Play it as a hard challenge, not as IQ training.
What can affect your Dual N-Back score?
N, most of all: each step back is one more pair of items to hold and update. Sound matters too, since a quiet speaker or an unclear voice makes letters easy to miss, and so does your strategy: saying the last few letters under your breath helps the Sound stream but can crowd out the squares. For memory under a different kind of load, try the Operation Span Test, which mixes letters with sums.
What doesn't the Dual N-Back Test measure?
It measures how well you hold and update two streams over about a minute, not intelligence, and it isn't a clinical assessment. Training versions adjust N from run to run as you improve; this one keeps the N you choose, so a score is for that N only.