ABA Fundamentals

Behavior of humans in variable-interval schedules of reinforcement.

Bradshaw et al. (1976) · Journal of the experimental analysis of behavior 1976
★ The Verdict

On two VI schedules, adults distribute button presses in the same ratio as the payouts, but the effect disappears when you add rules, VR components, or just test different people.

✓ Read this if BCBAs designing concurrent reinforcement programs in clinics or labs.
✗ Skip if Practitioners working with single-schedule DRA or FT tokens only.

01Research in Context

01

What this study did

Adults pressed a button for money on two VI schedules running at the same time. The researchers paid out on one schedule every 30 seconds on average, and on the other every 60 seconds on average.

They watched how fast people pressed and checked if the response split matched the payout split.

02

What they found

People pressed faster when the schedule paid more often. When both schedules were on, the button-press ratio lined up with the money ratio. The data fit Herrnstein’s matching equation.

03

How this fits with other research

Smith et al. (1975) ran almost the same setup and saw the opposite: people did not match. The difference is tiny. Both labs used adult humans and concurrent VI schedules. The 1975 paper simply shows matching is fragile, not absent.

Szempruch et al. (1993) later asked 30 adults to describe their own rules. Only 13 matched. The rest split their time evenly, favored one side too much, or picked only one button. Their self-talk predicted the pattern. The 1976 study did not collect rules, so it caught the ‘matching’ subgroup.

Winett et al. (1991) added a VR schedule to the mix. After brief training, people shifted from matching to maximizing. The 1976 result is therefore the baseline: pure VI schedules give you clean matching unless you add extra contingencies.

04

Why it matters

If you run concurrent reinforcement at home or in clinic, do not assume clients will ‘naturally’ match. Ask them what they think the rules are. A short think-aloud can tell you if they will split time evenly, chase the richer side, or lock onto one option. When you need balanced responding, keep the schedules simple and equal, and skip extra training that nudges toward maximizing.

FREE CEUs

Get CEUs on This Topic — Free

The ABA Clubhouse has 60+ on-demand CEUs including ethics, supervision, and clinical topics like this one. Plus a new live CEU every Wednesday.

60+ on-demand CEUs (ethics, supervision, general)
New live CEU every Wednesday
Community of 500+ BCBAs
100% free to join
Join The ABA Clubhouse — Free →
→ Action — try this Monday

Before starting a concurrent schedule, ask the client to say out loud how they think the rewards work and record their answer.

02At a glance

Intervention
other
Design
single case other
Population
neurotypical
Finding
not reported

03Original abstract

During Phase I, human subjects pressed a button for monetary reinforcement in five variable-interval schedules, each of which specified a different frequency of reinforcement. The rate of responding was an increasing, negatively accelerated function of reinforcement frequency; the data conformed closely to Herrnstein's equation. During Phase II, the same five schedules were in operation, but in addition a concurrent variable-interval schedule (B) was introduced, responses on which were always reinforced at the same frequency. Response rate in component A increased while the response rate in B decreased, as a function of the reinforcement frequency in component A. Relative response rates in the two component schedules matched the relative frequencies of reinforcement. Comparing the absolute response rates in component A during Phase I and Phase II it was found that introduction of the concurrent schedule did not affect the value of the theoretical maximum response rate, but did increase the value of the reinforcement frequency needed to obtain any particular submaximal response rate.

Journal of the experimental analysis of behavior, 1976 · doi:10.1901/jeab.1976.26-135