Rapid generation of balanced trial distributions for discrimination learning procedures: a technical note.

McGonigle et al. (2014) wrote a short how-to note. They give free computer code that spits out trial lists for any two-choice discrimination task.

The lists keep the left and right sides even. They also stop long runs of wins or losses that can sway choices.

The authors did not run new data. They simply show the code works and is ready to download.

Guilhardi et al. (2017) took the same idea into a classroom. Kids with autism liked lessons that mixed unknown and known trials more than big blocks of one kind.

Silverman et al. (1994) also cared about clean measurement. Their drug-disc task used point fines to curb random guesses, just like J et al. use balanced order to curb position guesses.

Blough (1980) warned that accuracy scores can lie if hidden biases steer the learner. The new code answers that warning by removing two big biases before the first trial starts.

Next time you set up a matching-to-sample or conditional-discrimination probe, paste the code into Excel. In five seconds you get a list that protects your data from side bias and win-streak bias. No extra math, no grant money, just cleaner sessions on Monday morning.