Work out how shape, count and fill change along the rows and columns of a 3×3 grid, then fill the empty cell. Items are generated from the rules on every run, so there is nothing to memorise.
Each of the three attributes gets its own rule, and those rules fix a single value for the empty cell. Every distractor changes exactly one attribute, so it necessarily breaks a rule.
Later items add more of the shifting-cycle rules; early ones hold a value constant within a row or column.
No. There is no norm sample and no reliability or validity study behind it. Treat the score as comparable only within this item format.
Yes. Rules and values are drawn at random and the grid is built on the spot, so repeats are very unlikely and a leaked answer key is worthless.
No. The task is whether you find the rule, not how fast you click, so nothing is timed.