Whit dae an n-back test actually measure?

The nummer shows the lag in the sequence. It isnae juist the total nummer o items a body can haud in mind, as the task needs ye tae keep watchin, updatin an response tae. Owen an colleagues' brain-imagin review includes monie versions that differ in stimulus identity, place, an verbal or nonverbal content. Owen et al., 2005

Whit for is it cried a workin-memory task

Theories o workin memory are aboot keepin information tae haun while ye uise or update it. The n-back procedure draws on pairt o that broad group o processes. The name disnae mean ilka workin-memory task measures the same ability or uises the same mix o storage an control. Baddeley, 2012 Cowan, 2001

Complex-span tasks, for ensample, pair rememberin information wi anither processin need an uisually ask for recall. Their methods review describes a different task structure frae continuous n-back recognition. A shared label cannae tak the place o evidence aboot hou scores relate. Conway et al., 2005

Can n-back tak the place o a span test?

Redick an Lindsey's meta-analysis fand weak correlations atween n-back an complex span, wi differences atween studies. Its conclusion is that thir measures cannae be uised interchangeably. Sae a weak score on ane an a strang score on the ither isnae automatically a contradiction or evidence that ane result maun be wrang. Redick et al., 2013

Accuracy depends on mair nor the lag

Stimulus duration, response deadline, target frequency an hou false alarms are dealt wi aw change the difficulty. Ayeways tell the difference atween detectin matches an pressin aften. A score needs its rule for baith richt detections an wrang responses, no juist a n level on its ain. The variety o methods in n-back research is ane reason levels fae different versions cannae be treated as universal norms. Owen et al., 2005 Redick et al., 2013

Comparisons by age an sex likewise need the same procedure an a suitable sample. Thir reviews dinnae gie a universal guid level by age for an online adaptive gemme. Settlin ane wad add a precision that the evidence disnae support. Redick et al., 2013 Conway et al., 2005

Whit dae better results wi practice show?

The n-back trainin meta-analysis distinguishes transfer tae related tasks frae mair distant results. The wider review o workin-memory trainin finds nae convincin general far-transfer benefit wi suitable comparisons. Sae gettin better at n-back is evidence o learnin the skill that wis tested, no a proved increase in general intelligence. Soveri et al., 2017 Melby-Lervåg et al., 2016

Reliability is a separate need. A task can engage a meaningful process an show a reliable group effect, but still no gie a steady eneuch rank for an individual. The findings on the reliability paradox warn against treatin a short gemme score as a fixed personal trait. Hedge et al., 2017

For watchin yer ain progress, keep the version, timin an scorin consistent, an track errors alangside richt responses. Gin ye change the stimulus types or move tae an adaptive version, note that ye hae changed the measurement. A clear description o the task tells ye mair nor a claim that ane nummer stands for aw workin memory. Owen et al., 2005 Redick et al., 2013