Anonymous sample measurement
We count sample entry selections, sample starts and completions, report-preview opens, generated-document preview use, document selection, page-picker, thumbnail and keyboard navigation, source/original/edited comparison selections, wording-diff views, fictional-test entry selections, invitation-page opens, invitation-form starts, submit attempts, save failures and exits without confirmation, saved invitation requests, document-search use and match navigation, outline and packet-citation selections, help use, and authenticated fictional-test starts. Evidence-finding task outcomes require the separate checkbox in that exercise as well as measurement being enabled. Search phrases, matched text, chosen source pages and issue-category values are not sent. First-test milestone, processing failure category, browser-reported export failure, and selected-feedback totals require a separate opt-in inside the workspace; the private progress needed to resume remains linked to your account. Each public sample event is counted at most once per tab, including navigation methods; these counts do not measure every click or prove that a page was read. A saved request is counted only once on the server; an authenticated test start is counted once per enrolled person and case when measurement is allowed at that moment. Exiting before a confirmation does not prove that a request was abandoned or not saved. These are separate event totals, not unique people or conversion rates. Sample events include no account, contact, record text, page URL, or referrer. Public viewing-position storage is separate functionality scoped to this tab and can be reset in the document preview. Browser privacy signals and your choice below disable sample and optional first-test measurement.