Teacher as token dispenser: Effect of an observer

Data were collected on rate per minute of administering token reinforcement for one male and six female teachers enrolled in a behavior modification course. Data were obtained from record cards maintained by pupils for tokens received during an observer-present condition and for an observer-absent condition. Comparison of reinforcement rates for observer-present vs. absent indicated significantly higher rates of token delivery (p <.025) in the observer-present condition. The observer effect has important implications for those programs in which assessment procedures introduce an observer into the classroom to collect data on changes in teacher behavior.