ESTES' STIMULUS SAMPLING THEORY. = statistical-learning theory/model = stochastic learning theory. The American psychologist William Kaye Estes (1919- ) formulated a mathematical learning theory that seeks to predict the exact numerical details of experimental results. The term mathematical learning theory denotes a type of approach to theory construction rather than a single, specific set of postulates that could technically be called a theory. Estes developed a form of mathematical learning theory in the 1950s called stimulus sampling theory (SST). SST started as a form of stimulus-response (S-

R) associationism that assumed that organisms learn by attaching new adaptive behaviors to stimulus situations where they formerly had inappropriate behaviors (cf., S-S learning model/theory - a learning approach that focuses on the association between stimuli, including both intervening variables and cognitive structures). Estes accepted E. L. Thorn-dike's empirical law of effect (i.e., reinforcers strengthen and guide behavior), although he does not subscribe to the "satisfaction" or "drive-reduction" properties of rewards. In SST, learning and performance are treated explicitly as a probabilistic or stochastic process (i.e., as a sequence of events that can be analyzed in terms of probability). The main dependent variable of statistical learning theory is the probability of various responses of a participant at any point in time (within a given learning theory), and a statistical learning model consists of assumptions about how the participant's probability of a correct response changes from trial to trial as a result of the outcomes experienced on each trial. In SST, the stimulus situation is represented as a population of independent variable components and the total environment ("stimulus elements"). At any given moment, only a sample of elements from the total population is effective or active, where the less variable the experimental conditions, the less variable are the successive trial samples of stimulus elements. The assumption of SST concerning responses is that their probabilities are determined by the proportions of stimulus elements in the sample connected to the various responses. The early experimental work in SST employed the probability-learning paradigm where the participant's task was to predict on each trial which of two events was going to occur; after the predictive response was made, the actual event was shown. Events in these probability learning experiments occurred in a random sequence with no information available to help in predicting perfectly which event will occur. The phenomena of forgetting and spontaneous recovery were interpreted by Estes in terms of random changes in factors in the stimulating environment from one experimental session to the next (e.g., factors such as temperature, humidity, participants' receptor sensitivity and attitudes). Estes' fluctuation theory of stimulus change accounts for the shapes of forgetting and recovery curves; it has been applied, also, to the phenomena of retroactive and proactive inhibition and verbal short-term memory. SST considers stimulus generalization in a manner similar to Thorn-dike's identical elements theory: a response associated with a stimulus population will generalize to a test stimulus to the extent that the second population shares common stimulus elements with the first population. Concerning discrimination learning, SST adopts the concept of selective attention and its associative relevant cues to help explain behavioral outcomes. Estes indicates that different learning models follow from SST when a small number of stimulus elements is assumed. Such "small-element models" fit the experimental data as well as do the original large-element models. Recent developments in Estes ' theory have changed in a direction closer to cognitive psychology and away from his original Guthrian stimulus-response approach. For example, Estes deals with the issue of participants' decision-making in preferential choice situations through his scanning model, which provides a viable approach to a process theory of decision-making. Estes also developed a hierarchical associations theory of memory that compares favorably with the "duplex" ideas of British associationism and with the "higher-order memory nodes" in J. Anderson and G. Bower's theory of memory. Although SST today has relatively few adherents as a "total" theory, the basic ideas of SST have been assimilated into a common stock of useful theoretical constructs. To date, Estes' SST is probably the most significant and rational attempt at a global quantitative learning theory in psychology. See also ASSOCIATION, LAWS/PRINCIPLES OF; ATTENTION, LAWS/PRINCIPLES/THEORIES OF; DECISION-MAKING THEORIES; DISCRIMINATION/GENERALIZATION HYPOTHESIS; FORGETTING AND MEMORY, THEORIES OF; GENERALIZATION, PRINCIPLES OF; GUTHRIE'S THEORY OF BEHAVIOR; HULL'S LEARNING THEORY; IDENTICAL ELEMENTS THEORY; INTERFERENCE THEORIES OF FORGETTING; PROBABILITY THEORY/ LAWS; THORNDIKE'S LAW OF EFFECT.

The 10 Keys To Happiness

The 10 Keys To Happiness

Simple Ways to Be Happy and Enjoy Your Life Regardless of Circumstances. Happiness is the underlying foundation that influences the quality of life. Have you ever seen someone who lives in a small house and has an older car? They may not be rich in terms of material things, but they are beyond rich in their happiness.

Get My Free Ebook

Post a comment