An observation on stimulus control in a tilt discrimination by children.

Kids learned to pick the one line that was tilted differently from three others. The task looked simple, but the researcher wanted to see what the kids actually learned.

Some children figured out an "odd-one-out" rule. Others memorized the exact tilt angle. The study tracked which strategy each child used.

Kids who used the oddity rule could pick the different line every time. Yet they could not say which direction the line was tilted. They learned the relation, not the feature.

Kids who skipped the oddity rule learned the exact tilt. They could name the angle later. The oddity shortcut blocked learning the specific stimulus.

Bloomfield (1967) first showed peak shift with tilted lines in pigeons. The birds' responding peaked away from the trained angle. Grusec (1968) asked whether children show the same shift or learn a simpler rule instead.

Wanchisen et al. (1989) later saw the same split in monkeys. First the animals used absolute tilt, then shifted to a relational rule when new tilts appeared weekly. The monkey data extend the child finding: control can swing back and forth.

Feinstein et al. (1988) frame the lesson for clinicians. Build the stimulus control you want into the original training. Do not wait to add generalization later. Grusec (1968) is a clear example: once the oddity rule took over, the absolute tilt was lost.

Check what your learner really knows. A child who always picks the red card may be using "not green" rather than "red." Test the exact feature you want to teach. Mix in novel samples. If the skill falls apart, oddity or another relation is in charge. Re-teach with the specific stimulus until it controls the response.