Measuring efficiency: trials to criterion, cost-benefit, and training time

Two teaching procedures can both work and still differ a lot in what they cost to get there. Efficiency measures tell you how much teaching, time, or effort it took to reach a goal, so you can pick the procedure that gets the learner there with the least waste.

Why it matters

Effectiveness asks "did it work?" Efficiency asks "how much did it take?" Instructional time is limited for every learner. A procedure that reaches mastery in half the sessions frees time for other goals, reduces exposure to errors and frustration, and makes better use of staff and family resources. As a BCBA, you will compare prompting strategies, decide between one-to-one and group formats, and plan how to train staff and caregivers. Efficiency data help you make those calls with evidence instead of habit.

On the exam, expect items that ask you to name the efficiency measure in a scenario, compare two procedures using trials or time, and spot comparisons that are not fair.

Key ideas

• Efficiency always refers to a criterion. You measure how many trials, how much time, or how many resources it took to reach a stated mastery criterion, such as 90% correct across two consecutive sessions. Without a fixed criterion, the numbers mean nothing.
• Trials to criterion counts opportunities. It is the number of response opportunities a learner needs to meet the criterion. Cooper, Heron, and Heward group it with percentage as a derived measure, because it is computed from counts of responding across teaching rather than being a basic dimension of behavior.
• Time to criterion counts instructional time. Also called training time or training duration, it is the total time spent teaching until the criterion is met. Sessions to criterion is a coarser version.
• Trials and time can disagree. A procedure with fewer trials can still take longer if each trial is longer, for example because of long prompting sequences, longer reinforcer access, or longer intertrial intervals. Look at both when trial length differs.
• Cost-benefit thinking weighs resources against outcomes. Resources include staff hours, training, materials, and family effort. Outcomes include skills mastered, maintenance, and generalization. A cheaper option is only better if outcomes are comparable.
• Efficiency can change across targets. When trials to criterion drops across successive sets of similar targets, the learner is acquiring new skills more efficiently, often described as learning to learn.

Definitions

• Trials to criterion: the number of response opportunities needed to reach a preset performance criterion.
• Time to criterion (training time): the total instructional time needed to reach the criterion.
• Sessions to criterion: the number of teaching sessions needed to reach the criterion.
• Mastery criterion: the predetermined level of performance that defines when a skill has been acquired.
• Cost-benefit analysis: comparing the resources a procedure requires with the outcomes it produces.

Examples from a session

• Comparing prompting procedures in a clinic. Ines, a BCBA, teaches Theo, age 4, to label animals. One set is taught with most-to-least prompting and a matched set with a progressive time delay. Theo meets the same criterion after 54 trials with time delay and 90 trials with most-to-least. Because trial length was similar, time delay was the more efficient procedure for Theo.
• When time tells a different story. At school, Mr. Osei compares two spelling procedures for Lina, age 11. Procedure A needs 30 trials at about 40 seconds each (20 minutes). Procedure B needs 45 trials at about 20 seconds each (15 minutes). Procedure A wins on trials, but Procedure B reached criterion in less instructional time.
• Learning to learn at home. Kenji, age 7, learns household safety signs in sets of five. Trials to criterion falls from 70 for set 1 to 40 for set 2 and 25 for set 3, with the same criterion each time. His acquisition is becoming more efficient.
• Cost-benefit in staff training. A clinic trains new RBTs to implement discrete-trial teaching. One-on-one coaching takes 6 supervisor hours per RBT. A video-model package plus brief feedback takes 2 hours, and treatment integrity reaches the same criterion. The video-based package is more efficient, as long as integrity data keep confirming it.
• Caregiver training. Rosa, a parent of a 15-year-old, learns to prompt a morning routine. Her BCBA tracks training sessions to criterion so the family's time is used well.

Common mistakes / exam traps

• Comparing across different criteria. Trials to 80% across two sessions and trials to 90% across three sessions are not comparable. Hold the criterion constant.
• Comparing unmatched targets. If one procedure gets easier targets, fewer trials may reflect the targets, not the procedure. Match targets for difficulty.
• Ignoring trial length. Fewer trials does not always mean less time. Check time to criterion when trials differ in length.
• Treating efficiency as the only goal. A fast procedure that produces poor maintenance or generalization is not a bargain. Efficiency matters only among procedures that are effective.
• Letting cost override client welfare. Cost-benefit thinking helps you use resources wisely, but a cheaper option that produces worse outcomes for the client is not the better choice.

Quick check

Related topics

• Measuring occurrence: count, rate, and percentage (C.3)
• Measuring time-based dimensions: duration, latency, and IRT (C.4)
• Choosing a measure that captures the critical dimension within real-world limits (C.9)
• Graphing data: line graphs, bar graphs, and cumulative records (C.10)
• Measuring procedural integrity: accuracy, dosage, and practical limits (C.12)