Mixed schedule

Components alternate without distinct signals

A mixed schedule contains at least two schedule components that occur successively. The active component changes according to a programmed rule, and no component-correlated stimulus tells the person which one is active.

One component might be variable-interval reinforcement and another extinction. Another arrangement might alternate two different ratio schedules. The component sequence can be planned while remaining unsignaled to the participant.

Multiple schedules add correlated stimuli

A multiple schedule also alternates components, but each has a distinct stimulus. A blue screen might signal one schedule and a white screen another.

That signal can support differential responding. Removing it changes the programmed arrangement from multiple to mixed, though leftover timing, staff, location, or material cues may still signal the component.

Unintended cues matter

A nominally mixed schedule can become discriminable when components always occur at the same clock time, one staff member delivers one component, or transitions follow a fixed pattern.

Audit these features before interpreting similar response rates as failure to discriminate or persistent behavior as schedule resistance. Record people, setting, materials, sequence, time, and recent outcomes.

Compare other compound schedules

In a chained schedule, distinct stimuli signal ordered components that culminate in terminal reinforcement. In a tandem schedule, ordered components occur without distinct signals. Concurrent schedules make alternatives available at the same time.

The BACB BCBA Test Content Outline includes distinguishing concurrent, multiple, mixed, and chained schedules. The outline is examination content rather than a treatment recipe.

A fictional example

A research team arranges alternating three-minute components for a button-press task. One component provides reinforcement on variable interval 30 seconds; the other provides none. The same screen, room, materials, and staff remain present.

Across four reinforcement components, Dara makes 44 responses in 12 minutes. Across four extinction components, Dara makes 38 responses in 12 minutes. Report rates of 3.67 and 3.17 per minute plus component order and transitions.

The modest difference does not prove that the schedule was truly indiscriminable. Recent reinforcement, carryover, fixed alternation, or subtle cues may contribute.

Keep component states in the data system

The person may receive no signal, while the implementation record must identify the active component. Timestamp every transition and programmed outcome.

Track responses, opportunities, reinforcement, obtained rate, latency, pauses, errors, and adverse effects by component. A pooled session rate erases the comparison that defines the schedule.

Transition effects deserve separate review

Responding can carry over from one component to the next. The first minute after a transition may differ from later minutes.

Predefine whether analysis uses the full component or separates transition windows. Avoid discarding early data after seeing a pattern.

Clinical usefulness is limited and specific

Mixed schedules can help analyze behavior when component signals are absent or study persistence under changing contingencies. They are less useful when a person needs clear predictability to make informed choices or navigate a routine.

Do not remove helpful schedules, communication, warnings, or transition cues simply to create a mixed schedule. Transparency, access, and client preference may outweigh a research-like arrangement.

An implementation checklist

• Which schedules form the components?
• How are order and duration determined?
• Which cues are intentionally absent?
• Which accidental cues will be audited?
• How is the active state logged?
• What measure is calculated for each component?
• How are transition periods handled?
• What would justify adding clear signals or stopping the arrangement?

Interpret the actual programmed and obtained contingencies, not the shorthand label.

Audit whether the components are truly unsignaled

List every feature that could differ when the schedule changes: staff posture, timer sounds, device latency, material position, room, instruction wording, clock time, recent reinforcement, and component order. Ask a second observer who cannot see the program state to predict the active component from the recorded session. Above-chance prediction suggests an accidental cue worth investigating.

Then compare programmed and obtained schedules. A variable-interval component that rarely produces reinforcement because sessions are short may function differently from the design. Record scheduled opportunities, actual outcomes, and interruptions. If an equipment delay always occurs in one component, treat it as part of the obtained arrangement.

Use a component-by-session table rather than one total. Keep missing or aborted components visible with reasons, especially when their absence changes the alternation pattern.

Replicate the sequence before treating one session's component difference as a stable effect.