Generalization

Generalization is the transfer of training effects from one situation to another. There are three types of student performance that may be considered evidence of generalization. The first, stimulus generalization, occurs when skills taught in one situation transfer to other situations. For example, a student who is taught by his teacher to say “please” when asking for assistance may begin to say “please” at the family dinner table when requesting another serving of food. In this example, the student uses a new skill (saying please) with different individuals (the family) in a new setting (the home) to make a new type of request (food).

A second form of generalization, response generalization, occurs when a strategy applied to one type of behavior changes a related behavior to which the strategy was not applied. For example, a program designed to reduce physical aggression in the school may also result in a reduction in other problematic behaviors such as truancy and verbal assaults.

Finally, maintenance, a third form of generalization, occurs when trained responses continue to be demonstrated across time after the completion of training. For example, a child who participates in social skills training to improve peer interactions may continue to use the trained skills with peers long after the formal training is terminated.

The goal of education is to teach students skills that will benefit them in other educational environments, in the community, and at home. However, generalization rarely occurs spontaneously. Therefore, it is [Page 223]important for educational experiences to be designed to promote the transfer of training effects.

One strategy for enhancing generalization is to make training conditions as similar as possible to natural conditions. For example, students learning about the value and exchange of money would benefit from classroom activities that simulate those under which money is exchanged and from field trips (e.g., going to the grocery store) with supervised opportunities for money to be exchanged. A similar strategy involves the inclusion of multiple examples in training. For example, when teaching students to label dogs, one might expose children to live dogs, pictures of several breeds of dogs, and abstract representations of dogs. Exposing students to multiple examples increases the likelihood that student performance will generalize to other examples and decreases the likelihood that performance will be limited to the one or two examples presented during training.

Teaching children to self-instruct and self-monitor are other ways to support generalization. For example, teaching children to state rules that describe the desired behavior will increase the probability that those skills will be performed in the teacher's absence. If students in a social studies class are taught the self-instruction, “Take out a sheet of paper and a pencil when the bell rings,” they are more likely to state the rule in math class and follow it.

When student performance depends on intensive training or management programs, generalization to nontraining conditions may not occur. To enhance generalization, training programs should be designed so that when the desired behavior is performed, it is reinforced naturally in the environment. For example, a special education student with severely limited communication skills might be taught to operate a voice-output device to request attention from others. Although intensive training may be required to perform the behavior, the natural responses of others may be sufficient to reinforce and maintain the behavior.

...