Improving accuracy of portion-size estimations through a stimulus equivalence paradigm.

Koegel et al. (2014) taught college students to judge food portions without cups or scales. They used a stimulus-equivalence program: first train exact matches (picture of one cup rice → word 'one cup'), then test if new pairs emerge (word 'one cup' → different rice photo).

Sessions ran in a quiet lab. No extra apps or tools were allowed. The team checked if the new skill held one week later and worked with brand-new foods.

After equivalence training, students eyeballed portions almost perfectly. Accuracy stayed high one week later. Most learners also judged novel foods they had never trained on.

The study found positive results.

Rehfeldt et al. (2005) tried a similar equivalence plan with adults who had developmental disabilities. They linked tastes to pictures and found those links lasted 30–40 days. Koegel et al. (2014) now shows the same logic works for everyday nutrition skills in neurotypical adults.

Mazur et al. (1992) warned that visual-only training often fails to generalize. Koegel et al. (2014) added clear verbal labels during training, matching E’s advice and getting the broad transfer that visual-only studies missed.

McKeel et al. (2017) later used the PEAK system to build taste-picture-word classes in adults with autism. Their success backs up L’s core idea: once equivalence classes form, people can apply them to new items without extra teaching.

If you teach portion control to clients, you can borrow the same quick equivalence steps: train a few exact matches with real photos and labels, then test for emergence. The skill can maintain for at least a week and spread to new foods, cutting the need for repeated trials or measuring cups. Try it next session with one common food and one label; check if the learner gets untrained photos right.