The Front Desk: Judgment When Everything Arrives at OnceThe Front Desk: Judgment When Everything Arrives at Once. A service desk does not fail because somebody is slow. It fails in the minute when four things want the same person.

Methodology: Measures desk judgment when more work arrives than one person can hold, through original situational, selection and self-report items keyed to published constructs. Construct statement: It measures what a person does when more work arrives than one person can hold at once at a service desk: what they do first, what they hand to somebody else, what they park with a promise, and what they decline without losing the person in front of them. It does not measure speed, typing, memory, product knowledge, language fluency, or how friendly a person sounds. Declared response instruction: behavioural tendency throughout - every situational stem asks what the respondent is most likely to do rather than what a person should do, because instructed-knowledge framing is the more fakeable of the two. Item format: an interrupt queue authored as a FIXED STATE. A static item bank cannot branch, so the state of the desk at every arrival is authored and is identical for every respondent, and each arrival is scored against the moves available at that arrival rather than against a position the respondent created. Sixteen arrivals run in four blocks of four; each arrival offers exactly four moves, one of which does the work now, one hands it to another person, one parks it with a promise, and one declines it, and every option carries a graded partial-credit weight and an authored base rate. Full credit sits on doing it now in five of the sixteen arrivals, on handing off in four, on parking with a promise in four, and on declining in three, so a respondent who takes everything on cannot score well. Four recovery items place the respondent in a moment already gone wrong; four selection items ask what would be left for the next person at the end of a shift and are graded by set overlap, which punishes leaving too much and too little alike; and eight balanced-keyed self-check statements ask about the respondent's own habits, half worded so that agreement is not the flattering answer. Scoring is graded situational judgment, chance-corrected against the declared per-option base rates before any rung is assigned, because a graded partial-credit percentage has a floor near half by construction and would otherwise place almost every respondent in the middle band. The composite carries its standard error of measurement at both the sixty-eight and ninety-five per cent spans, and the reliability behind those spans is a stated target rather than a measurement until enough sittings exist. No expert panel has yet rated these items, so the rung boundaries are design targets and are described that way on the report. Each of the four capabilities is reported as a three-way classification and never as a number or a percentile, because eight or fewer exercises cannot support one. Construct areas drawn on: research on triage and queue discipline in service settings, and the difference between first-in-first-out handling and handling by consequence of delay; the service-recovery paradox and the evidence that the second failure costs more than the first; work on waiting-line psychology, in which unexplained and unoccupied waits feel longer than explained ones; delegation research on what makes a task transferable and the costs of retaining work that somebody else could do; handover research from shift-based work, where a promise survives a change of staff only if it moves with its detail; the literature on emotional labour and surface acting at the front line; research on customer incivility and boundary-setting by frontline staff; work on psychological safety and asking for help before a queue builds; the goal-setting and implementation-intention literature on committing to a time rather than to an intention; and research on interruption and task-resumption cost in interrupt-driven work. All items are original works, no trademarked instrument or branded methodology is named or reproduced, and no affiliation with any source is claimed. AssessAll original design.