What Cognify is
Revelian developed Cognify, guided by the Cattell-Horn-Carroll theory of cognitive abilities, and launched it in late 2016. Criteria acquired Revelian in 2020 and now offers it. Criteria’s help centre describes it as a game-based assessment of cognitive aptitude, the ability to acquire, retain and use new information, made up of mini-games that together give one overall measure.
The standard version is three games, Grid Lock, Numbubbles and Proof It, and takes about 10 minutes not counting tutorials. A six-game version adds Resemble, Short Cuts and Tally Up and takes about 25 minutes; employers have to request it. UCognify is a language-independent version with no verbal game, in two-game and five-game forms.
The three standard games
Grid Lock measures problem solving: you fit all the pieces into a grid as quickly as you can. Criteria’s information brief gives it 9 rounds of increasing difficulty with a 3-minute limit.
Numbubbles measures numerical reasoning: you pop the bubbles whose equation equals the target number. The brief gives 10 rounds with a total limit of just under 3 minutes.
Proof It measures verbal knowledge: you identify misspelled words and punctuation errors. The brief gives 5 rounds in 5 minutes.
Criteria publishes no description of Resemble, Short Cuts or Tally Up beyond their names, so if your invitation lists six games, three of them cannot be previewed from Criteria’s own material.
What taking it is like
Criteria says the games are fast-paced and you may not use a calculator. You get real-time feedback during play, round features are randomised, and puzzles are streamed at random from sets of equal difficulty, so candidates do not all see the same items.
Cognify runs on any device. Criteria’s research found no significant difference between mobile and desktop scores in a sample of 280, and it recommends an up-to-date device and browser and a reliable connection. Its candidate guide notes that on a laptop or desktop you will need a mouse.
Revelian’s older help pages said candidates could choose the order of the games. Criteria’s current guidance is that each game must be completed in the order presented.
How it is scored
Each game is scored on several metrics, such as time to respond, as well as on whether you were right. Game scores are compared with a norm group to give percentiles, and the overall score is a weighted average of the game scores, with weights that depend on the version.
The report shows an overall percentile and sub-scores for Problem Solving, Numerical Reasoning and Verbal Knowledge. Criteria’s score report guide calls the overall percentile the most important figure, and bands results as Below Average (0 to 19), Average (20 to 79) and Above Average (80 to 100).
How our practice versions work
We have practice versions of all six games, played separately: Grid Lock (8 boards in 8 minutes), Numbubbles (8 rounds in 5 minutes, 12 bubbles a round, each on a hidden timer and shaking before it bursts), Proof It (8 passages in 6 minutes, with the number of errors shown), Short Cuts (10 puzzles in 8 minutes), Resemble (15 figures in 10 minutes) and Tally Up (12 rounds in 6 minutes).
Our clocks are longer and our round counts different from the ones Criteria publishes: the real Grid Lock is 9 rounds in 3 minutes, Numbubbles 10 rounds in just under 3, and Proof It 5 rounds in 5. Short Cuts, Resemble and Tally Up are our reconstructions, since Criteria does not describe them: finding the most efficient route to a target, rebuilding a rotated figure on a grid, and comparing two groups of value tokens.