Cued Continuous Performance Test (CPT-OX)

One letter at a time in the centre of the screen, 200 ms each, one every 1850 ms, with a blank screen in between and no fixation mark. All letters have the same colour. The participant taps anywhere on a large response area. A version with 12 shapes (circle = cue, diagonal cross = target) is available.

Useful when a study needs sustained attention together with response preparation and inhibition. The O prepares a response: the participant taps when an X follows it (Go) and withholds the prepared response when another letter follows it (NoGo). Go and NoGo trials are equally frequent, so they can be compared without a frequency (oddball) effect. The task is used in ADHD, addiction and psychiatric research, often with EEG.

Not a diagnostic instrument. Not a quick EMA probe: the published task has 400 stimuli, about 12 minutes. A single 100-letter block gives too few errors to compare individuals on omission or commission counts (10 Go and 10 NoGo trials), and healthy adults make close to no errors.

Keep the letters, the practice, the default timing and the 400 stimuli of the published task (Nash et al., 2013), about 12 minutes plus practice. For short repeated sessions in which hit RT and its variability are the main outcomes, one 100-letter block (about 3 minutes) is available: the published 400 letters consist of four identical blocks of 100 (McLoughlin et al., 2010), but no published study reports a single block on its own.

• Stimuli: the letters A B C D E F G H J L O X with O = cue and X = target (default; Nash et al., 2013), or 12 shapes with circle = cue and diagonal cross = target (an adaptation without a published source)
• Main block of 400 stimuli (default; Nash et al., 2013) or 100 (one of the four 100-letter blocks of McLoughlin et al., 2010), always 20% cues, 10% Go, 10% NoGo, 10% X without a preceding O and 50% other letters
• Stimulus duration (200 ms) and blank (1650 ms), as in Nash et al. (2013) and Fallgatter et al. (2013)
• Practice of 24 stimuli with 3 Go and 2 NoGo trials (Tye et al., 2017) and a message after each error; runs once by default. Optional pass criterion, which has no published value: all Go trials hit with at most one commission, repeated up to a set number of rounds

• Hits, hit rate and omissions (n and %); taps later than 1500 ms count as omissions and are also reported as late taps
• Commission errors in total and per type: NoGo (O followed by another letter), cue, X without a preceding O, other letters
• Mean, median and SD of the hit RT, and the coefficient of variation (SD / mean)
• d′ for Go hits against NoGo commissions (log-linear correction) and anticipations (taps under 200 ms)
• Per stimulus: letter, condition, drawn onset and on-screen duration, every tap with its RT, and the score; the seed of the sequence

A hit is the first tap 200–1500 ms after a Go X (McLoughlin et al., 2010), and each tap belongs to the stimulus shown last before it. Every session draws a new pseudo-random sequence from a stored seed, under fixed rules: every O is followed by an X or by a NoGo letter, no X follows an X, at least one other letter separates two cue pairs, no more than three Go or NoGo pairs follow in a row, and both halves of the block have the same counts. Error counts are near floor in healthy adults (about 0.5 omissions and 0.4 commissions in 400 stimuli; Nash et al., 2013), so when error rates are the outcome, use the 400 stimuli or pool sessions. The practice has the length and Go/NoGo counts of Tye et al. (2017); its 2 lone X's and 12 other letters are not published and keep the main block's proportions. No published study gives a numeric practice criterion (“comprehension ascertained”; McLoughlin et al., 2010), so practice runs once unless the optional criterion is turned on. In shape mode the circle is the cue, whereas in Go/No-Go shape mode the circle is the go stimulus; if both tasks run in one study, the circle means different things, which is one reason letters are the default. Onsets are taken after the screen has been drawn and taps at touch-down, but this timing has not yet been validated on real devices.

• The cued CPT with the letters O and X comes from Fallgatter's group, which used it to measure the NoGo-anteriorisation of the P300 (Fallgatter, Brandeis & Strik, 1997). As described by Nash et al. (2013), it shows 400 letters for 200 ms each, with a 1650 ms interval.
• The task has been used to study response suppression in alcohol dependence (Fallgatter et al., 1998) and response inhibition as a predictor of self-control (Nash et al., 2013).
• The CPT-OX of the ADHD literature uses the same proportions (10% Go, 10% NoGo, 10% X without a preceding O), 400 letters in four identical blocks of 100, and scores hits 200–1500 ms after the target (McLoughlin et al., 2010).
• Every accessible study in both lines of work uses 400 stimuli, including a remote administration on home computers (Fallgatter et al., 2013; Kaiser et al., 2021; Sun et al., 2023).
• The paradigm descends from the original continuous performance test, in which participants respond to an X only when it follows an A (Rosvold et al., 1956).

• Fallgatter, Brandeis & Strik (1997), NoGo-anteriorisation in a cued CPT
• Fallgatter et al. (1998), response suppression in alcoholics
• Nash et al. (2013), response inhibition in a cued Go/NoGo task predicts self-control
• McLoughlin et al. (2010), the CPT-OX in adult ADHD
• Fallgatter et al. (2013), the cued CPT with 400 letters and an 1850 ms SOA
• Tye et al. (2017), a 24-trial practice block with three Go and two NoGo sequences
• Rosvold et al. (1956), the original continuous performance test
• Hautus (1995), corrections for extreme proportions in d′

• Go/No-Go
• Stop-Signal Task
• RSVP Target Detection