The effects of varying the distribution of generalization stimuli within a constant range upon the bisection of a sound-intensity interval by rats.

Raslear (1975) taught rats to pick the middle loudness between 50 and 90 dB tones. After training, the team ran generalization probes with extra tones spaced in different ways.

The probe tones stayed inside the same 50–90 dB window, but the gaps between them changed. The goal was to see if the spacing moved the animals’ idea of the midpoint.

When probe tones were packed closer together, the rats’ bisection point slid toward the dense end of the range. Wider spacing pulled the midpoint the other way.

The loudness range never changed—only the spacing of test tones did. This shows stimulus spacing alone can bend the gradient even when the true limits stay fixed.

Stevenson (1966) saw the same spacing effect in pigeons pecking colors. Closer test stimuli flattened the gradient and cut peak shift. The 1975 rat study extends that rule to hearing, showing the effect holds across senses and species.

Nevin (1969) used the same bisection task with brightness and found no context shift. The rat data now contradict that calm picture—spacing did matter for sounds. The clash warns us that vision and hearing may follow different context rules.

Schroeder et al. (1969) showed that a 24 h delay also flattens gradients unless you add a quick warm-up. Taken together, both delay and spacing can warp generalization probes, so you need to guard against both when you plan stimulus sets.

If you probe generalization after teaching a discrimination, the gaps between your test items can nudge where the learner places the boundary. Pack too many examples near one end and the child may call that end the middle. Space probes evenly and across the full range to keep the gradient honest. Check your stimulus sheet before the next session—equal steps protect your data.