Key evidence & numbers
| Task | Required comparison | What a high score means directly |
|---|---|---|
| 1-back | Current item versus previous item | Good performance under that rule |
| 2-back | Current item versus two positions earlier | Good updating and matching for that task |
| 3-back | Current item versus three positions earlier | Performance at a larger lag, under specified timing |
The number indicates the lag in the sequence. It is not simply the total number of items a person can store, because the task also requires ongoing monitoring, updating and responding. Owen and colleagues' neuroimaging review includes multiple versions differing in stimulus identity, location and verbal or nonverbal content. Owen et al., 2005
Why it is called a working-memory task
Working-memory theories concern keeping information available while using or updating it. The n-back procedure engages part of that broad family of processes. The label does not mean every working-memory task measures exactly the same ability or uses the same mix of storage and control. Baddeley, 2012 Cowan, 2001
Complex-span tasks, for example, combine remembering information with another processing requirement and generally ask for recall. Their methodological review describes a different task structure from continuous n-back recognition. A shared label cannot replace evidence about how scores relate. Conway et al., 2005
Can n-back replace a span test?
Redick and Lindsey's meta-analysis found weak correlations between n-back and complex span, with heterogeneity between studies. Its conclusion is that these measures cannot be used interchangeably. A weak score on one and a strong score on the other is therefore not automatically a contradiction or evidence that one result must be false. Redick et al., 2013
What affects it?
Accuracy depends on more than the lag
Stimulus duration, response deadline, target frequency and the handling of false alarms change difficulty. Always distinguish detecting matches from pressing frequently. A score needs its rule for both correct detections and incorrect responses, rather than an isolated n level. The methodological diversity in n-back research is one reason that levels from different implementations cannot be treated as universal norms. Owen et al., 2005 Redick et al., 2013
Age and sex comparisons likewise require the same procedure and a suitable sample. These reviews do not provide a universal good level by age for an online adaptive game. Assigning one would add precision that the evidence does not support. Redick et al., 2013 Conway et al., 2005
What does improvement with practice show?
The n-back training meta-analysis distinguishes transfer to related tasks from more distant outcomes. The broader working-memory training review finds no convincing general far-transfer benefit under appropriate comparisons. Improving at n-back is therefore evidence of learning the tested skill, not a demonstrated increase in general intelligence. Soveri et al., 2017 Melby-Lervåg et al., 2016
Reliability is a separate requirement. A task can engage a meaningful process and show a reliable group effect without supplying a sufficiently stable individual ranking. The reliability-paradox findings caution against treating a short game score as a fixed personal trait. Hedge et al., 2017
For self-observation, keep the version, timing and scoring consistent and track errors alongside correct responses. If you change stimulus types or move to an adaptive version, note that you changed the measurement. A clear description of the task is more informative than a claim that one number represents all working memory. Owen et al., 2005 Redick et al., 2013
Try the related free test
N-back test
Try the n-back test to experience updating and matching items at a specified lag. Read its scoring instructions and compare only equivalent runs. The result is a score on this task, not an IQ diagnosis or a universal memory capacity.
Common questions
Does 3-back mean I can remember only three things?
No. The number specifies the comparison lag, while monitoring and updating contribute to performance. Owen et al., 2005
Is n-back the same as digit span?
No. The meta-analysis finds that span and n-back tasks are not interchangeable working-memory measures. Redick et al., 2013
Does getting better prove my IQ increased?
The training literature does not establish that inference. Task learning and broad transfer must be evaluated separately. Soveri et al., 2017 Melby-Lervåg et al., 2016
Sources
- Owen et al. (2005). N‐back working memory paradigm: A meta‐analysis of normative functional neuroimaging studies.
- Redick et al. (2013). Complex span and n-back measures of working memory: A meta-analysis.
- Cowan (2001). The magical number 4 in short-term memory: A reconsideration of mental storage capacity.
- Baddeley (2012). Working Memory: Theories, Models, and Controversies.
- Conway et al. (2005). Working memory span tasks: A methodological review and user’s guide.
- Soveri et al. (2017). Working memory training revisited: A multi-level meta-analysis of n-back training studies.
- Melby-Lervåg et al. (2016). Working Memory Training Does Not Improve Performance on Measures of Intelligence or Other Measures of “Far Transfer”.
- Hedge et al. (2017). The reliability paradox: Why robust cognitive tasks do not produce reliable individual differences.
Explore your daily routines in Helibrain
The app has short tests of reaction time, working memory, self-control, reasoning and creativity. Retake them whenever you like, see your trend, and find habits that may help.
Start free in your browser
