Replicating stimulus‐presentation orders in discrimination training

The team taught receptive labels to children with autism. They tried two orders of showing pictures and spoken words.

One order showed the sample picture first, then the choice pictures. The other order flipped that sequence.

They used an alternating-treatments design. Each child got both orders in mixed blocks within the same session.

Both orders worked for most kids. Sample-first and comparisons-first each led to mastery.

Adding extra repeats of the same order did not help or hurt. The child’s own speed was the only clear difference.

Wong et al. (2020) also tested order, but looked at free rotation of pictures instead of fixed left-to-right. Their free-rotation won. The two studies line up: small tweaks in how we show items can change speed, yet kids still learn.

Koegel et al. (2014) asked whether to start with simple-only trials or jump straight to conditional arrays. They found skipping simple trials saved time. Bergmann’s team now shows that, once you run conditional trials, the exact cue order is flexible. Together the papers say: simplify the task sequence first, then pick the order the learner likes.

O’Neill et al. (2018) and O’Neill et al. (2022) changed prompt timing, not stimulus order. They also used alternating treatments and got clearer winners (progressive prompt delay). The contrast reminds us some variables matter more than others; prompt delay beats fixed delays, but sample-first versus comparisons-first is largely a tie.

You can stop worrying about the “perfect” order. Run a quick probe: present the sample first for ten trials, then the choices first for ten. Track correct responses. Keep the sequence that gives the learner the higher score right away. No extra repetition is needed, so sessions stay short and efficient.