Primary stimulus generalization and secondary extinction as a function of strength of conditioning.

Primary stimulus generalization may be ;nned as the tendency for stimuli other than it CS, but falling within the same stimulus mension as the CS, to elicit responses when aining has been only to the CS. Pavlov's rly work (1927) in classical conditioning dicates that geneialization first increases, en decreases, with the number of training ials (reinforcements). These results have :en substantiated by more recent research Jovland, 1937b; Thompson, 1958). Two measures of generalization are used. "ie first is absolute generalization, which is ken simply as the response-eliciting strength a given test stimulus, recorded usually as •quency or amplitude of response. The cond type, relative generalization, is taken the response eliciting strength of a test imulus as compared with the strength of R to the CS. Relative generalization would en seem to be the more logical measure to .e when the level of acquisition is varied, ice the strength of CR to the CS is taken to account, thus permitting more precise mparisons between levels of conditioning or dining. Associated with stimulus generalization is condary extinction, which is a decrement in rength of a CR due to extinction of a similar sponse (e.g., response to a generalization st stimulus). Secondary extinction may be ought of as generalization of extinction. It is been found (Bass & Hull, 1934; Hovland, J37a) that the effectiveness of a generalization p area as an extinguisher decreases with the [stance of the test area from the original CS ttining area.