A demonstration of higher‐order response class development in children

Dixon et al. (2018) worked with two children. One had autism. One was neurotypical.

The team used matching-to-sample to teach a higher-order class. Kids learned to unscramble mixed-up sight words.

After training, the children saw brand-new words. Ninety-eight percent of these words had never been taught.

Both kids could unscramble the new words right away. They did it in single words and in full sentences.

No extra teaching was needed. The higher-order class let them generalize for free.

Paliliunas et al. (2022) later used the same logic with kids with autism. They taught hierarchical categories instead of word play. Emergent responding still showed up.

Noell et al. (2026) is a near twin. They ran discrete trials to build hierarchical classes in preschoolers. Kids got naming and transformation of function without direct teaching.

Hayashi et al. (2013) did an earlier literacy demo. After kids learned to select letters when they heard the name, most started naming letters on their own. All three studies show the same pattern: brief equivalence-style training creates new skills you never drilled.

You can stop mass-teaching every single word. Build a higher-order class with a few examples and let generalization do the rest. Try it for spelling lists, sight words, or even braille. One short protocol can save hours of drill and still give you novel, accurate responding.