Glossary

The shared vocabulary for this course. Every lesson uses these terms exactly as defined here. Skim it once; return to it whenever a word feels slippery.

① Cognitive domains — what gets impaired

Cognitive domain

A separable mental capacity (attention, memory, speed, control). "Impairment" is never one global dial — it is degradation across specific domains, and different causes hit different ones. DSM-5 / Sachdev et al. 2014

Attention domain

The capacity to select and sustain processing of relevant information. Splits into sub-types below.

Sustained attention / vigilance

Keeping focus steady over time on a monotonous task. The first thing fatigue erodes; measured by lapses, not averages. Alhola & Polo-Kantola 2007

Divided attention

Handling two streams at once (e.g., track a target while responding to a cue). The first thing low-dose alcohol erodes. PLOS ONE 2013

Selective attention

Focusing on one signal while ignoring a competing distractor (the basis of the Stroop and Flanker tasks).

Processing speed / psychomotor speed

How fast you take in, decide, and physically respond. Broadly sensitive to almost any impairment, which makes it a good general screen but a poor diagnostician of cause. Jaeger 2018

Working memory

Holding and manipulating information for seconds (e.g., remember a sequence, then act on it). Degraded by alcohol, cannabis, and benzodiazepines.

Executive function

Top-down control of thought and action. Per the field-standard Miyake model it is not one thing but three separable components below — so it needs more than one task to measure. Miyake et al. 2000

— Inhibition

Suppressing a prepotent (automatic) response. The Go/No-Go and Stroop tasks probe this; alcohol degrades it markedly.

— Updating

Continuously refreshing working-memory contents (the n-back task probes this).

— Shifting

Switching flexibly between rules or tasks (set-shifting; the Trail Making Test B probes this).

Time perception

Accurate sense of elapsed time. Distorted distinctively by cannabis — a near-signature, since most other impairers don't hit it the same way. Figueiredo et al. 2021

Impairment fingerprint

(Course term.) The characteristic pattern of which domains a given cause degrades, by how much, and over what time course. The reason a multi-task battery can sometimes tell what kind of impairment, not just whether.

Wake-state instability

The signature of sleep loss: performance doesn't slow uniformly — it becomes erratic, with brief lapses interleaved among normal responses. Van Dongen et al. 2003

② Tasks — how we probe the domains

Reaction time (RT)

Time from stimulus to response. Simple RT: one stimulus, one response. Choice RT: pick the right response among several (adds decision load). The atomic unit of most impairment tasks.

Psychomotor Vigilance Task (PVT) gold standard

Respond as fast as possible to a stimulus appearing at random intervals over several minutes. The most validated impairment measure because it is reliable, sensitive to sleep loss, and — crucially — has almost no learning curve. Dinges & Powell 1985; Basner & Dinges 2011

PVT-B

A brief ~3-minute PVT designed for repeated/field use. Retains much sensitivity, but shortening trades off some validity. Basner et al. 2011; Frontiers 2022

Lapse

A reaction time above a threshold (commonly ≥500 ms on the PVT) — a momentary failure of attention. Counting lapses is often more sensitive than averaging RT. Basner & Dinges 2011

Stroop task

Name the ink colour of a colour word (e.g., "RED" in blue ink). Probes selective attention + inhibition via response conflict.

Flanker task

Respond to a central arrow while flanking arrows point the same or opposite way. Probes selective attention + inhibition. (NIH Toolbox uses this.)

Go/No-Go

Respond to frequent "Go" stimuli but withhold on rare "No-Go" ones. Probes response inhibition; errors of commission = inhibition failure.

n-back

Report whether the current item matches the one n steps back. Probes working-memory updating; load scales with n.

Trail Making Test (TMT)

Connect dots in order. A: numbers only (speed/scanning). B: alternate numbers/letters (adds set-shifting). Varjacic et al. 2022

Digit Symbol Substitution Test (DSST)

Match symbols to digits against the clock using a key. A broad processing-speed screen — highly sensitive, low specificity. Jaeger 2018

③ Measurement — deciding a result is trustworthy

Reliability

Consistency of a measure. Test-retest reliability (target ICC > 0.8): does the same un-impaired person score about the same on Tuesday as Monday? Without this, you can't tell impairment from noise. Akshoomoff et al. 2014; JMIR 2025

Validity

Whether the measure captures what you claim. Construct/concurrent validity: does it agree with an established measure of the same thing?

Sensitivity

Of truly impaired people, the fraction the test flags (true-positive rate). High sensitivity = few impaired people slip through.

Specificity

Of truly un-impaired people, the fraction the test clears (true-negative rate). High specificity = few sober people falsely flagged.

Baseline (individual) vs. norms (population)

Compare a score against the same person's sober history (baseline) or against a population average (norms). Baselines cut false positives in naturally slow people and false negatives in naturally fast ones — generally the stronger design. Baseline study 2013

Reliable Change Index (RCI)

The statistic that decides whether today's change from baseline exceeds normal test-retest noise (accounting for measurement error and practice). The principled "impaired vs. just an off day" rule. UDS RCI 2016

Practice / learning effect

Scores improving simply from repeated exposure, not from any real change. A confound for repeated testing — it can mask impairment. The PVT is prized for resisting it.

Ceiling effect

A task so easy that even impaired people score near-perfect, leaving no room to detect degradation. The silent killer of a "game" that's too fun and too easy.

Fitness-for-duty testing

Performance-based testing that measures the effect (impaired function) rather than the cause (a substance). Catches fatigue, illness, and medication that a breathalyzer misses. Burns & Hiller-Sturmhöfel 1995