Training letter discrimination in four-year-old children.

Researchers worked with 24 four-year-olds who had no reading skills. They used a match-to-sample game to teach letter discrimination.

Kids picked the correct letter card after seeing a sample. Some kids earned tokens only for choosing letters with the right features. Others got tokens for any choice. A third group played with no tokens.

Children who were reinforced for critical letter features made a large share fewer errors on the final test. The other two groups showed no improvement over baseline.

The critical-features group also learned faster, needing fewer trials to reach mastery.

Frederiksen et al. (1978) later showed that errorless prompting beats trial-and-error during reversal learning in older kids. Both studies agree: reduce errors early, don't wait for mistakes.

Barlow et al. (1973) tested a different errorless method—gradually showing the wrong letter. They also cut errors to near zero in preschoolers. The 1972 study adds that you can achieve the same goal by simply reinforcing the right features from day one.

Kim et al. (2023) now supersedes the 1972 setup by showing that shorter, 12-trial mini-sessions with individual mastery criteria teach sight words faster. The lesson: keep the discrete-trial format, but tighten mastery rules and session length.

When you run letter or shape discrimination programs, deliver praise, tokens, or stickers only when the learner points to the defining part—like the diagonal in 'K' or the circle in 'O'. Skip reinforcement if the child picks based on color or size. This single shift cuts errors and training time, freeing you to move to harder skills sooner.