Harver Assessment Practice: Every Module, Scored the Harver Way

Cognitive and chat results: 0 to 100%, where 50% is average. Harver says most people score between 30% and 70%. That is not a percentile, where only 40% of people would fall in that range; it is a norm scale centred on the average candidate. Each cognitive subtest compares you with a norm group named in the module, often an education level such as Bachelors, and the score depends on how many items you get right.

The SJT: a percentage, with a penalty for reversals. Each best or worst pick earns +1 when it matches the key, -1 when you mark the keyed best as worst or the keyed worst as best, and 0 for a neutral option. Harver's own example report labels an 88% retail SJT a Great Fit.

Typing: net words per minute. Typed words minus mistyped words, divided by time, compared with Harver's published average of 38 and standard deviation of 17. Because Harver publishes its norm, our typing feedback places you on Harver's own numbers.

Personality Print: no better or worse. Harver's report guide says high scores are not necessarily better and low scores are not necessarily bad. Candidates who get a feedback report see each trait in one of three bands, with no numbers, and our feedback uses the same format.

The match decides. None of those is the decision. The employer screens on an overall matching score that combines the modules against the vacancy's profile, with its own Good Fit and Great Fit brackets. On a standard job profile each cognitive subtest is roughly a tenth of that match. No pass mark is published, and whether you see your own results is the employer's choice.

Harver's knowledge base calls its cognitive module "NOA Cognitive Ability", and its vacancy builder prints the type "NOA" under each subtest. NOA is the Amsterdam publisher of the MCT capacity tests, and Harver's five subtests are five of NOA's eight item types, with NOA's names in brackets. The education level in a module name, such as "(Bachelors)", is NOA's way of norming.

The difference is the report. NOA's own candidates see a stanine drawn as Lager, Midden or Hoger; Harver's see the 0 to 100% scale. So the same performance reads differently on the two, and our Harver games report in Harver's terms. If you have been sent NOA's own test instead, see the NOA practice tests.

Harver publishes no per-subtest clocks. We use NOA's published figures for the same item types, except two: Spatial Ability gets 10 minutes, because a Harver screenshot shows 9:20 left on its first item, and Perceptual Speed gets 5, to sit inside Harver's stated 5 to 15 minutes. Our Perceptual Speed score is correct minus wrong, so random tapping cannot pay; Harver does not publish its rule.

The chat timings are ours. Harver publishes the 12 minutes and the three customers, but not when the second and third arrive, how many messages each sends, or how its four scores are weighted. Ours brings customers in at 0:40 and 2:30, or sooner once you have resolved everyone else, and weights the four equally.

The SJT and the typing text are ours. Harver's standard SJTs are for contact centre, retail and team lead roles and are not public, so our scenarios are written for the same settings and competencies and keyed on Harver's published scoring rule. Harver's default typing text is not public either; ours mixes plain sentences with the addresses, numbers and dollar sums its screenshots show, with a four-minute cap.

No practice rounds. The real cognitive subtests and the chat simulation open with a practice step. Ours go straight to the scored items, and each game's instructions say so.

What we do not simulate. The Harver-LTI language test, which is a licensed third-party instrument; the spoken language test and the writing assessment, which we cannot score faithfully; job knowledge tests, which each employer writes; AI Prevail, whose format is unpublished; the content moderator assessment; the video interview; and the legacy HEXACO, Learning Agility and Cultural Fit modules. Personality Print Professional (146 questions, 22 traits) is not simulated; our game is the Hourly version.