ABA glossary
Interobserver agreement
Also called IOA
Definition
The degree to which two or more observers who independently measure the same behavior report the same values, usually expressed as a percentage. Many teams treat about 80% as a minimum.
In plain words
IOA is a check on whether the data can be trusted. Two people record the same session without looking at each other's sheets, and their records are compared. High agreement suggests the operational definition is clear and the observers apply it consistently. Low agreement means someone, or the definition, needs attention before anyone makes decisions from the data. IOA does not prove the data are correct: two observers who misunderstand the definition in the same way can agree perfectly.
Example from a session
The BCBA and the RBT both count Ana's requests during the same 30-minute session, each on their own sheet. The RBT records 18 and the BCBA records 20. Total count IOA is the smaller count divided by the larger count, times 100: 18 ÷ 20 × 100 = 90%. The RBT records her own data honestly rather than adjusting it to match the BCBA's count.
For BCBA candidates and supervisors
Analysts choose an IOA method that fits the measure and is strict enough for the decision: total count, mean count-per-interval, or exact count-per-interval IOA for counts; total duration or mean duration-per-occurrence IOA for duration; trial-by-trial IOA for discrete trials; and interval-by-interval, scored-interval, or unscored-interval IOA for interval data (scored-interval IOA is stricter for low-rate behavior, unscored-interval for high-rate). A common practice is to collect IOA in at least 20% of sessions within each condition. Low IOA calls for retraining, clarifying the definition, and checking for observer drift.
Often confused with accuracy, reliability
IOA compares one observer with another; accuracy compares the record with what really happened. High IOA raises confidence in the data but does not establish accuracy. Reliability means repeated measurement of the same events gives the same result. IOA is sometimes loosely called "interobserver reliability," but it measures agreement between observers, which is not the same thing.
Exam tip
Know the basic formulas: total count IOA = smaller ÷ larger × 100; interval-by-interval IOA = agreements ÷ (agreements + disagreements) × 100. When an option says high IOA "proves" the data are accurate, it is wrong.
Can you spot interobserver agreement in a scenario?
Practice RBT A.8 with exam-style questions and instant feedback.