First contribution to study the development of analogical reasoning with eye-tracking. Adults follow a projection-first strategy (study A:B, then C:D). Children organize their search around C.

No evidence of alignment (i.e., A-C and B-D saccades) in the A:B:C:D paradigm.

Children have difficulties to inhibit the main goal of the task (“what goes with C”).

Errors result from a less efficient strategy starting at the beginning of the trial.