Sustained Attention and Monitoring Assessment for Control Room, Security and Quality Monitoring WorkNothing happens for forty minutes. Then something does.

Methodology: Thirty-eight original exercises across five formats: ten drawn tracking exercises in which cued objects become indistinguishable and move, twelve keyed single-choice exercises, five select-every-that-applies exercises, five binary claims and six self-report ratings on a six-point scale with the keying balanced. Three response instructions are declared and each strand is reported apart from the others. The tracking strand is a performance task - what you were still holding when the movement stopped. The knowledge strand asks what is most likely to be true or most effective, which is a knowledge instruction. The self-report strand asks what is like you, which is a behavioural-tendency instruction. None of the three is averaged into another, because a capacity, a piece of knowledge and a self-description are three different things and a single number over them would describe nobody. Construct statement: it measures how many moving things somebody can hold at once, what they know about how a long watch fails and how to build one that fails less, and how accurately they read their own attention. It does not measure intelligence, conscientiousness, reliability, safety or fitness for any role, it is not a clinical instrument and it diagnoses nothing. Scoring is a one-tailed risk scale. A tracking result above the printed threshold produces no finding at all, which is deliberate: the construct only matters at one end, so reporting the whole range as a level would invite a reading the measurement cannot support. Below the threshold the report prints a flag with a described behaviour bound to a situation, never a percentile, and always paired with what a watch can be changed to. The flat-baseline gate is run on the continuous tracking index that feeds the risk scale rather than on the risk scale itself, because a one-tailed scale is zero for most respondents by construction. Every corrected figure is corrected against an authored answer prior per exercise rather than against a uniform random draw, because people do not answer at random: on a counting exercise they undercount, and the priors say so. Those priors are a model and the report says so; no percentile appears anywhere and none could honestly be claimed until real sittings replace them. Reliability is reported as McDonald's omega estimated from the item count and a stated assumed inter-item correlation, in the open, because no live sample exists; a strand whose omega will not carry a number is given a three-way placement instead and the refusal is printed. Constructs and sources: the vigilance decrement and its early onset (Mackworth 1948; Teichner 1974); resource-depletion accounts of sustained attention and the cost of a concurrent task (Warm, Parasuraman and Matthews 2008); signal probability and the criterion shift that rarity produces (Wolfe, Horowitz and Kenner 2005; Green and Swets 1966); multiple-object tracking capacity of roughly four items and its independence from effort (Pylyshyn and Storm 1988; Scholl 2009); individual differences in tracking capacity and their stability (Oksama and Hyona 2004); redundancy and the shared-miss effect in paired monitoring (Bahrami et al. 2010); alarm false-alarm rate and the cry-wolf effect on response latency (Breznitz 1984; Bliss 2003); rest-break scheduling and the recovery of detection (Ariga and Lleras 2011); automation function allocation and the cost of monitoring an automated process (Parasuraman and Riley 1997; Bainbridge 1983); metacognitive blindness to one's own attentional lapses and mind-wandering (Schooler et al. 2011); time-on-task feedback and self-regulated breaks (Ross et al. 2014). All items are original works written for this instrument. No commercial attention instrument, continuous performance test, scale name or report layout is reproduced, referenced or implied, and no affiliation with any test publisher exists or is claimed.