A letter n-back for working-memory updating: decide for every letter whether it matches the one n positions earlier. Defaults follow Kane et al. (2007): 2-back and 3-back.
Configure parameters and run an interactive preview exactly as participants will experience it. No data is recorded.
Adjust parameters below, then start the preview on the right.
Include practice trials
Shown with feedback before the main task
Task parameters
8 blocks (2-back, 3-back, 2-back, 3-back, 2-back, 3-back, 2-back, 3-back), 384 items, about 20 min, plus about 5 min of practice. Each block has 48 items, 8 of which are matches (never more than 2 in a row); the first n items of an n-back block cannot match yet and are not scored. Each item starts with a fixation cross; responses count from the item until the next fixation cross, and the screen is blank in between.
New tasks start with the settings of Kane et al. (2007): 2-back and 3-back alternating, 8 blocks of 48 letters (B, F, H, K, M, Q, R, X, in random upper or lower case), 8 matches per block, a 500 ms fixation cross, the letter for 500 ms and a 2000 ms blank, after one 40-item practice block per load. That is about 25 min with practice. Shorter settings, such as fewer or shorter blocks, are lab choices, not published values. The load of the first block is not stated in the paper (we start with 2-back, as in its practice order); its lure sequences and the rule that each letter appears 6 times per block, once as a match, are not reproduced.
Outputs per load: hits, misses, false alarms, correct rejections, hit and false-alarm rates, d′ (log-linear correction), median RT of hits and accuracy; plus the load cost, highest minus lowest load, in accuracy and median RT.
This is a researcher preview. No participant data is recorded.
Simulated participant view
A letter n-back for working-memory updating: decide for every letter whether it matches the one n positions earlier. Defaults follow Kane et al. (2007): 2-back and 3-back.
No data is recorded
A fixation cross, then a letter in the centre and a blank interval; the participant taps Match or No match for every letter, until the next fixation cross. Letters appear in random upper or lower case. A cue screen before each block gives its rule. Digits or shapes can replace the letters.
Useful at a baseline or lab session when a letter n-back with Match / No match responses and signal-detection scores is wanted, as in Kane et al. (2007).
With the published settings the task takes about 25 minutes, too long for repeated momentary assessment; the digit N-back is the shorter default. Less suitable when a working-memory-capacity estimate is the goal: n-back scores correlate only modestly with complex-span tasks.
New tasks start with the settings of Kane et al. (2007): 2-back and 3-back alternating, 8 blocks of 48 letters (4 per load) from B, F, H, K, M, Q, R and X in random upper or lower case, 8 matches per block, a 500 ms fixation cross, the letter for 500 ms and a 2000 ms blank, after one 40-letter practice block per load. Shorter settings, such as fewer or shorter blocks or a 0-back baseline, are lab choices; report them as such. The paper does not say which load comes first; the app starts with 2-back, the order of its practice blocks.
With the defaults, the load cost compares 3-back with 2-back. A 0-back block asks only for detection of a named target and can serve as an attention and motor baseline. d′ uses the log-linear correction (Hautus, 1995), which keeps perfect and zero rates finite; Kane et al. (2007) computed sensitivity and bias with the logistic formulas of Snodgrass and Corwin (1988), so the values are not directly comparable. The first n items of each block cannot match and are not scored. Kane et al. also built lures (foils repeating a recent letter at another distance) into their lists in a controlled way; the app's random sequences do not reproduce them, and avoid immediate repeats except as 1-back matches.
Add N-Back (letters) to your ESM study on SMAAT — free to get started.