Resistance to change of operant variation and repetition.

Firth et al. (2001) worked with three pigeons in a small lab chamber.

The birds pecked three keys in a row.

Sometimes they had to vary the pattern. Other times they had to repeat the same pattern.

After each pattern, food arrived. The birds learned both rules.

Next, the team added mild shocks, skipped meals, or gave free food. They counted how many patterns the birds still finished.

When the birds had to vary their key order, they kept working through every shock or skipped meal.

When they had to repeat the same order, they quit sooner.

Variation made the behavior tougher, even though both rules paid the same amount of food.

Singh et al. (1991) looks like a clash. They mixed new spelling words with old ones for kids with learning disabilities. The mix did not help. The pigeon study says variation strengthens behavior. The difference is the measure. H et al. counted how long behavior survives disruption. N et al. counted how fast new words were learned. Survival and learning are not the same thing.

Kay et al. (2020) and Galtress et al. (2012) used the same quick-swap design. They swapped prompts across trials to see which one wins today. H et al. swapped rules across trials to see which one lasts tomorrow. Same tool, different clock.

Muething et al. (2024) studied relapse after FCT. They asked, "Will the old behavior come back?" H et al. asked, "Will the current behavior stay?" Both questions live under the same roof: how behavior endures when the world shifts.

You can use variability as armor. When you teach a child to mand, mix the words, the tone, and the partner across trials. The response class will survive the busy classroom, the substitute teacher, or the day the AAC device dies. Start with solid reinforcement, then test with mild disruption like background noise or delayed praise. If the mand keeps going, you have built a tough, variable class.