Wetin n-back test dey really measure?

The number show the lag for the sequence. E no simply mean the total number of items wey person fit keep for mind, because the task still need person to monitor, update and respond as e dey go. Owen and the other researchers' brain-imaging review include different versions wey vary for the kind of stimulus, location, and whether the content na verbal or nonverbal. Owen et al., 2005

Why dem call am working-memory task

Working-memory theories talk about keeping information available as you dey use am or update am. The n-back method use some of this wide group of processes. The name no mean say every working-memory task measure exactly the same ability or use the same mix of storage and control. Baddeley, 2012 Cowan, 2001

For example, complex-span tasks join remembering information with another processing requirement, and dem usually ask you to remember and report the information. The review of their methods describe task structure wey different from continuous n-back recognition. Sharing one name no fit replace evidence about how the scores relate. Conway et al., 2005

N-back fit replace span test?

Redick and Lindsey meta-analysis find small correlations between n-back and complex span, and the studies no all get the same result. The conclusion be say you no fit use the two measures like say na the same thing. So, low score for one and high score for the other no automatically mean say dem contradict each other or one result must be wrong. Redick et al., 2013

Accuracy depend on more than the lag

How long stimulus show, response deadline, how often target appear, and how dem handle false alarms fit change the difficulty. Always separate noticing matches from pressing button often. To understand score, you need know the rule for both correct detections and wrong responses, no be just the n level by itself. The different methods researchers use for n-back na one reason why levels from different versions no fit count as universal standards. Owen et al., 2005 Redick et al., 2013

Comparing age and sex too need the same procedure and a suitable group of people. These reviews no provide one universal good level for each age for online adaptive game. If we assign one, e go make the evidence look more precise than e be. Redick et al., 2013 Conway et al., 2005

Wetin improvement with practice show?

The meta-analysis about n-back training separate transfer to related tasks from results wey dey farther away. The wider review of working-memory training no find convincing general far-transfer benefit when researchers use suitable comparisons. So, getting better for n-back show say you don learn the skill wey dem test; e no show say general intelligence don increase. Soveri et al., 2017 Melby-Lervåg et al., 2016

Reliability na another thing wey matter. Task fit use one meaningful process and show reliable effect for a group, but still no give stable enough ranking for each person. The reliability-paradox findings warn against treating short game score like fixed personal trait. Hedge et al., 2017

If you dey observe yourself, keep the version, timing and scoring the same, and track mistakes together with correct responses. If you change the stimulus types or switch to adaptive version, note say you don change the measurement. Clear description of the task dey tell more than claim say one number represent all working memory. Owen et al., 2005 Redick et al., 2013