An evaluation of a progressive high-probability instructional sequence combined with low-probability demand fading in the treatment of food selectivity.

Penrod et al. (2012) worked with two boys with autism who ate only a few foods. The team wanted to know if a gentle, step-by-step method could help them try new foods without force.

They started each meal with easy, high-probability tasks like touching or smelling the food. Then they slowly added harder, low-probability tasks such as taking a tiny bite. No escape extinction was used—kids could say “no” and still get praise for earlier steps.

Both boys began eating the new foods after a few weeks. Acceptance rose without tears, gagging, or restraint.

Parents and teachers said mealtimes felt calmer. The boys also kept eating the new foods when the team later checked back.

Sheppard et al. (2026) tried the high-probability part alone and saw mixed results—one child ate more, two did not. Their study shows the fading step Becky added may be the key piece.

Luiselli (2000) and Andersen et al. (2024) also blended demand fading with other supports and got good feeding gains. Together these papers build a line of evidence that gradual, antecedent-focused packages can work across diagnoses.

Older studies like Sisson et al. (1993) used parent-led hierarchical bites at home. Becky’s lab version keeps the same gentle spirit but adds the high-p warm-up, giving clinicians a clearer protocol to replicate.

You can copy this sequence tomorrow: start with touch, then smell, then a lick, then a bite while praising every small step. No need for physical prompts or escape extinction, so families who fear “force-feeding” are more willing to try. Track data each meal; if progress stalls, add the fading steps that Sheppard omitted before considering tougher tactics.