“Not established” and “passed” are not the same answer.
If no performance validity measure was given, the honest output is that the question could not be asked. Software that treats an absent gate as a cleared one has made an assumption on the clinician’s behalf, in the one place where assumptions do the most damage.
Three outcomes, not two
| Situation | What Casebook says |
|---|---|
| A validity measure was given and is above its cutoff | Asked, and nothing to report |
| A validity measure was given and is below its cutoff | Reported, with the cutoff and its source |
| None was given | Could not ask — drawn differently from the first row |
The third row is the point. “Asked and answered no” and “could not ask” are opposite facts, and a page that runs them together has said something it does not know. They are drawn differently throughout — a refusal gets a dashed rail rather than a clean one — because the distinction has to be visible at a glance and not buried in a caveat.
One failure is not a finding
A single below-cutoff result is a place to look, not a conclusion. Two or more says so, and hands it to the clinician without a conclusion attached.
Nothing is ever labeled invalid. That word is a judgment about a person’s engagement, sometimes with consequences for a claim or a case, and it is not a judgment that follows from a number. A failure carries the genuine reasons a real presentation fails one: severe impairment, pain, fatigue, medication, limited English, and a genuine effort problem all produce the same low score.
The cutoffs, and only where they are unambiguous
The catalog knows TOMM, WMT, MSVT, NV-MSVT, Reliable Digit Span and the CVLT’s forced-choice trial. Each cutoff is the published one, carried with its source — the TOMM manual (Tombaugh, 1996) for below 45; the standard WMT cutoff at 82.5% or lower. The CVLT-3’s forced-choice cutoff is the CVLT-II’s, carried over to the same 16-item trial, and it says so.
Where the literature disagrees on a cutoff, the score is judged at the more specific line and the disagreement is printed beside it. A failure only at the disputed line is shown and never counted as one of two. Where there is no line to choose between, the score is held and displayed without a verdict.
Why the gate runs first
Every reading after this one is conditional on it, so it cannot sit further down the page. An interpretation of a profile nobody has established is readable is an interpretation of noise — and it will read exactly like an interpretation of signal, which is the whole problem.
Casebook does not score and does not diagnose. It holds scores computed elsewhere and describes what the pattern is, with the arithmetic open underneath.
See a gate that refused
The demonstration includes cases where the question could not be asked, because a tool that only shows its confident output is not showing you much.