Improvement of procedural fidelity in discrete‐trial programs using computer‐based instruction to teach skilled observation

Lionello‐DeNolf et al. (2025) tested a new computer lesson called Train-to-Code. The goal was to help staff run discrete trials the right way every time.

Workers first tried a short set of trials on the computer. The program gave live cues and instant feedback. Later the same workers ran real trials with a child to see if the lesson stuck.

After the computer lesson, staff followed the trial steps much better. Errors dropped and good moves stayed high when they worked with real learners.

The gains showed up fast and held without extra coaching.

Machado et al. (2021) also used a computer to train observers. Their group cut scoring errors while watching fast videos. Both studies prove a short, automated module can lift fidelity, one for scoring and one for teaching.

Al-Nasser et al. (2019) and Hansard et al. (2018) got similar boosts with paper packets or solo videos. Train-to-Code adds dynamic prompts that change as you work, an update to those earlier static tools.

Spanoudis et al. (2011) warned that first-trial data can miss mistakes. The new program sidesteps this by logging every response during practice, matching that older caution with tighter measurement.

You can plug staff into Train-to-Code for thirty minutes and see cleaner trials the same day. No extra trainer time, no printouts. Use it during onboarding or as a quick tune-up before starting new programs.