Over-selectivity as a learned response.

McAleer et al. (2011) asked if over-selectivity is something we are born with or something we learn. They split neurotypical adults into two groups. One group practiced match-to-sample with simple cards that held three pictures. The other group practiced with busy cards that held nine pictures.

People who first trained with the nine-picture cards later ignored more cues on a brand-new task. Their over-selectivity was stronger. People who started with the three-picture cards paid attention to more cues later. The study showed that over-selectivity can be taught by the way we first set up training.

Farmer-Dougan et al. (1999) extends the idea to learners with intellectual disability. They proved that prompting clients to point to every part of a compound card cuts over-selectivity right away, but the gain fades when you drop the prompt.

Gomes-Ng et al. (2023) is a direct successor. They found that if a learner is already highly over-selective, later extinction of the favored cue helps the hidden cues gain control. If the learner is only mildly over-selective, the same extinction can actually hurt control by the weaker cues. The 2023 paper uses the 2011 learned-over-selectivity idea to decide who benefits from extinction.

Brayner de Freitas Gueiros et al. (2020) used the same kind of compound cards but with preschoolers and a Go/No-Go method. They did not test over-selectivity, yet their fast emergence of untaught relations shows the method still works with young kids.

When you build early discrimination programs, start with simple compound boards. Fewer pictures per card now means broader attention later. If you inherit a learner who is already highly over-selective, run a quick probe, then try extinction on the dominant cue only for the high-score cases. Keep prompts that force observing responses if you want the gain to stick.