Three-configuration matching-to-sample in the pigeon.

Researchers taught pigeons a matching game. The birds saw one shape on a center key and had to peck the same shape on a side key.

Only three of four possible side-key positions were used during training. Later the team added the fourth, never-seen position.

They wanted to see if the birds would still match correctly when the new spot appeared.

The pigeons bombed. Most birds started pecking the same physical spot again and again, even when the shape there was wrong.

Strong position bias took over. Accuracy crashed when the new configuration showed up.

Robinson et al. (1974) saw the same crash with new colors and shapes. Together the studies warn: pigeons (and maybe kids) lock onto the exact way you first set the task.

Sprague et al. (1984) looks like a contradiction at first. Their birds matched brand-new figures just fine. The fix was tiny: they let the pigeons see the side keys a little earlier, breaking the position habit.

Nelson et al. (1978) and Cook (2002) show the broader rule. When training mixes in non-matching trials and keeps the correct key consistent, pigeons learn a flexible rule instead of a rigid position.

Your learner might act like these pigeons. If you always flash the correct picture on the right, the child may just point right. Rotate seats, shuffle cards left and right, and mix in easy wrong choices early. A few extra examples during teaching beats big trouble during testing.