Stimulus sampling theory began as an attempt to formalize certain of Guthrie’s ideas. Various versions of stimulus sample theory are in vogue today, as will be briefly considered later, so in this sense Guthrie lives today. In the Estes (1950) version of stimulus sampling theory, the stimulus is represented as a population of elements. On a trial, only a sample of elements from the entire population is effective (i.e., sampled). Two sources affecting which elements are sampled are random events (noises, etc.) and changes in the participant (variable behavior over trials, etc.). One view of stimulus sampling suggests that each element has a particular probability of being sampled. Another is that a number of elements is drawn without replacement from the population elements. Still another view suggests other possibilities. Estes assumed that each element is connected to one response. In this manner different S-R associations would be formed.