Your last three hundred cases, available to the next one.
Casebook holds every score you have — from any platform, any instrument, any re-evaluation — as one record per person. Then it shows you what a new profile looks like against the series you have already built.
The demonstration is the real product, holding invented people. No sign-up, no card, nothing to uninstall.
The four hours after the testing
The scoring is done. What is left is six vendor portals, a PDF, a spreadsheet, and a score table you type by hand — and then the same eight sentences about it you wrote last week about somebody else.
And the part that actually needed your training, comparing this profile against the hundreds you have seen, happens entirely in your head, because the hundreds are in a filing cabinet.
What it does
Brings the scores in
Upload a finished report or a vendor export — Q-global, MHS Online, anything that prints a score table. It reads the tables, works out which instrument each section belongs to, picks up the informant and the age, and proposes a row per score. Nothing is written until you have looked at it.
Puts the battery on one axis
A WISC-V index, a CVLT trial and a BRIEF composite expressed in the same units, by domain, with a composite folded out when its parts are there so one testing is not counted twice. Ability and symptom scales both, each pointing the right way.
Reads it against your own series
How unusual is this shape among the people you see? Which of your cases looked like this, and what did you conclude about them? No published norm table answers that. Your own case series does, and you already have one.
Every statement opens to its arithmetic
Casebook describes patterns, and shows the rule it applied, the comparisons that rule turned on, the scores underneath, what it could not determine, and where the threshold was published. On paper the workings print with it.
It says so when it cannot tell. A question that could not be asked gets its own section, separate from the questions that were asked and answered no — because a report that runs those together has said something it does not know.
What it will not do
It does not score
Scoring belongs to the publishers who own the norms, and you already pay for it. A score reaches Casebook already computed, by whichever platform you licensed to compute it.
It does not diagnose
No output names a disorder or ranks the likelihood of one. That is enforced by a test that reads every sentence the software can produce and fails the build if it reaches for one.
It does not overstate
Ipsative comparison reports at the rarity line, not the significance line — a “significant” Wechsler split occurs in around 40% of healthy people. Subtest-level profile analysis was removed outright.
It holds one identifier, and only if you choose
A date of birth, when you enter one, is the single direct identifier Casebook keeps — the age at each test is computed from it. No name, no record number, no email, no phone, no address — in any table that holds patient data, and a test fails if a column with one of those names appears. A person is otherwise a code you choose, and the mapping from that code to a human being stays with you.
A report you upload carries all of it. The text is read for that one request, parsed, and kept nowhere: no column, no cache, no copy on disk. What is stored, exactly.
Built by someone who writes these reports
Casebook started as one clinician’s own instrument and grew into this. Every rule in it was argued with the literature first — three of the first four readings had to be weakened once the retest data was read properly, and those corrections are the part worth having.