Failure to Replicate the Effects of the High-Probability Instructional Sequence on Feeding in Children With Autism and Food Selectivity.

The team tried the high-probability instructional sequence (HPS) with three autistic children who only ate a few foods.

First they gave easy instructions the kids usually followed, like "touch nose." Right after each easy yes, they gave a feeding instruction, like "take bite of apple.

They tracked bites across meals to see if the sequence boosted food acceptance.

HPS did not help. Food acceptance stayed flat for all three children across meals.

The failure matches no change at all, not even a small upward blip.

Older papers summarized in Busch et al. (2010) once showed HPS helping picky eaters. Those studies were small and rarely used autistic samples.

Chezan et al. (2019) is a direct replication, so the clash is an apparent contradiction: earlier hints of success now disappear when the same steps are repeated with new autistic kids.

Northgrave et al. (2019) ran a different 2019 replication and found that letting kids pick reinforcers slowed teaching. Together the two 2019 papers warn that tiny procedural choices can make or break an intervention.

Do not assume HPS will fix feeding problems. Run a short probe with data before you build it into a treatment plan. If bites do not rise within a few sessions, pivot to other evidence-based feeding tactics like differential reinforcement or texture fading.