W. K. Estes’s research while affiliated with Rockefeller University and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (43)


Intelligence and Learning
  • Chapter

January 1981

·

13 Reads

·

14 Citations

W. K. Estes

Given the title of this volume, some of the questions one should expect to be at issue are surely: What has been, what is, and what should be the relationship between learning and intelligence? Are the referents of the two terms identical? Are they, rather, related like two sides of a coin? Or do they perhaps refer to levels of intellect or intellectual function?


Judgments of relative frequency in relation to shifts of event frequencies: Evidence for a limited-capacity model
  • Article
  • Publisher preview available

July 1979

·

4 Reads

·

16 Citations

Journal of Experimental Psychology Human Learning & Memory

In 2 experiments with 104 young adult Ss, shifts of item frequencies in a judgment-of-relative frequency paradigm were used to test differential implications of limited vs unlimited capacity models. In a within-list design, relative frequencies of presentation of items (words or CVC strings) were shifted after different numbers of observation trials, and tests were given after different numbers of postshift trials. During both pre- and postshift learning, judgments of relative frequency approached an asymptotic level of frequency matching, and the degree to which judgments were determined by memory for frequency proved relatively independent of number of preshift observation trials. These results are most compatible with a limited-capacity context model, which also provides a quantitative account of the full pattern of data over a limited range of observation–test cycles. (15 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)

View access options

Some Functions of Memory in Probability Learning and Choice Behavior

December 1976

·

6 Reads

·

41 Citations

Psychology of Learning and Motivation

Publisher Summary The chapter examines scale properties of human choice behavior. A little is known about the psychological process that gives rise to the scale properties. Scales involved in most choice and decision behavior of adult human beings must be generated by some as yet unidentified learning processes. The studies presented in this chapter represent a first step toward illuminating the problem. This chapter focuses on two types of data, one having to do with subjective probability of events, and the other with utility or reward value. In each case the first step demonstrates the orderly patterns of behavior reflecting scale properties that can be generated in controlled learning situations. The next step analyzes each situation from the standpoint of current theories of memory and attempt to bring out experimentally some specific aspects of storage and retrieval that are responsible for the development and maintenance of the orderly patterns of choice behavior associated with psychological scales.


The Cognitive Side of Probability Learning

January 1976

·

64 Reads

·

311 Citations

Psychological Review

Models proposed for probability learning largely represent performance rather than learning theory. Perhaps for this reason, quite different models have been required to provide accounts of data arising from different experimental paradigms. In the present approach, a common theoretical framework is sought in concepts of coding and organization in memory. Following the theoretical analysis, an observation-transfer paradigm is developed that permits the study of predictive behavior depending on categorical, as distinguished from episodic, memory. The paradigm was tested in 3 experiments with 116 Ss: Exp I assessed the effects of variation in both stimulus frequency and outcome probability, Exp II examined the role of stimulus familiarity, and Exp III attempted to vary the encoding of trial outcomes in memory. Results suggest that probability learning and transfer derive from frequency learning. The individual categorizes events and forms representations in memory of relative frequencies of event categories. When different cues in a multiple-cue, probability learning situation occur equally often, this process yields predictive behavior closely reflecting the probabilities that alternative events associated with a cue will occur when the cue is present. But when cue frequencies are unequal, the categorical memory model implies that predictive behavior may be grossly out of line with actual probabilities. In general, depending on task requirements, predictive responses are either direct reflections of relative frequency judgments or are governed by strategies involving an additional level of encoding of event categories. (59 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)


Table 3 Response Breakdown by Stimulus Position for Change vs, No-Change Conditions With Data Averaged Over Quadrants
Serial position functions for letter identification at brief and extended exposure durations

January 1976

·

111 Reads

·

185 Citations

Attention Perception & Psychophysics

The properties of serial position functions for tachistoscopic report were investigated over a wide range of viewing times. Four-letter strings of random consonants were presented in varying display locations relative to the fixation point with the observers’ eye movements monitored to limit them to a single fixation for each display. Salient properties of the serial position curves include an overall central-peripheral gradient, higher performance at the ends than the interior of letter strings regardless of absolute location, and left-right asymmetry in the visual field, all of these being largely independent of viewing time. Errors reflecting loss of positional information are prominent even at extended viewing times, are more nearly symmetrical in the left and right visual fields than other types of errors, and, in contrast to item errors, occur less frequently in letter sequences that have high frequencies in English. Further, transposition errors exhibit a pronounced peripheral-to-central drift, possibly reflecting gradients of positional uncertainty. Such gradients may be implicated in the peripheral-central asymmetry of the lateral interference effects exerted by other letters on a target letter in a nonfoveal location.


Some targets for mathematical psychology

August 1975

·

24 Reads

·

110 Citations

Journal of Mathematical Psychology

Concern among mathematical psychologists regarding the present status of the discipline arises from two sources. One is the growing disparity between the rapid development of mathematical psychology as an adjunct to research and the slower and more uncertain progress toward a cumulative body of theory. The other is the problem of adjusting to the encroachment of computers and computer simulation models into the traditional province of mathematical methods. It appears that these problems might be alleviated by recognition of the complementary aspects of mathematical and computer simulation approaches to psychological theory and by a shift of emphasis from tactics to strategy in the construction and evaluation of models of both types.


The locus of inferential and perceptual processes in letter identification

June 1975

·

19 Reads

·

171 Citations

Journal of Experimental Psychology General

Conducted 2 experiments with a total of 56 Ss which attempted to clarify the nature and temporal course of the processes responsible for the often reported superior identifiability of letters presented in a word context over letters embedded in nonwords or presented alone. Exp I confirmed and extended results of other studies concerning the effects of word or nonword context under a forced-choice detection procedure. Additional findings in Exp II suggest that the inferior performance on single letters is almost entirely attributable to a greater incidence of omission errors; however, the difference between words and nonwords is largely attributable to the decreased incidence of transposition errors in a word context. In general, the effects of the linguistic context in which a target letter is embedded depend strongly on the point in time at which the contextual information becomes available. Overall findings are discussed in terms of a theoretical "perceptual filter" model and the idea that Ss can report perceptions of a word without being aware (without showing in their behavior) that they have first identified the constituent letters. (29 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)



Learning theory and intelligence

October 1974

·

23 Reads

·

200 Citations

American Psychologist

Proposes that the primary objective of intelligence testing be redirected from predicting to improving intellectual performance. In a number of instances, specific behaviors involved in responding to items on intelligence tests have been subjected to experimental analysis in laboratory studies. Results support the concept that different levels of test performance result from different combinations of processes and mechanisms of coding, hierarchical association, memory search, and cybernetic control of intellectual functioning. (20 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)


Detection of single letters and letters in words with changing vs unchanging mask characters

March 1974

·

231 Reads

·

27 Citations

Bulletin of the Psychonomic Society

Using a forced-choice procedure and tachistoscopic displays, accuracy of identification was compared for letters presented alone vs the same letters embedded in words or in nonword letter strings. Ss were 12 young adults. The advantage for single letters over letters embedded in words or nonwords was found to depend jointly upon characteristics of the pre- and postmasks and change vs no change in mask characters in nontarget locations at the onset of single letter displays. (PsycINFO Database Record (c) 2012 APA, all rights reserved)


Citations (42)


... However, beyond the empirical interest and practical usefulness of a measuring tool, one of the aims of the scientific inquiry is to clarify what this tool precisely measures in order to better understand its diagnostic and prognostic power. The complexity of the processes underpinning the apparently simple behavior of immediately repeating a list of verbal items probably explains why simple spans have sometimes been considered as poorly diagnostic of school achievement and intelligence (e.g., Estes, 1981;Glasser & Zimmerman, 1967;Wechsler, 1958), some authors having even as-This document is copyrighted by the American Psychological Association or one of its allied publishers. ...

Reference:

Simple Spans Underestimate Verbal Working Memory Capacity
Intelligence and Learning
  • Citing Chapter
  • January 1981

... Like other exemplar models (Nosofsky, 1988), our 1010 similarity based model retains memories for each past trait instance, allowing it to be sensitive to 1011 frequency information in a way that is not available to RL models (Hackel & Amodio, 2018;1012 Rescorla & Wagner, 1972) that track an average value. Nonetheless, our task context is different 1013 from those in which RL models are often tested, as there was no explicit incentive or goal for 1014 participants to learn the group norms as they were merely instructed to self-endorse traits while 1016 RL models may fair better in contexts where there is an explicit goal of learning about the group 1017 (i.e., instrumental learning), as frequency-based models have previously performed better in 1018 observational contexts where learning is passive (Estes, 1976a(Estes, , 1976b Goldwater, 2021) during social learning. We hope that future work in social cognition research 1041 might implement computational models that incorporate normative relations and similarity 1042 structure, in addition to learning target values. ...

Some Functions of Memory in Probability Learning and Choice Behavior
  • Citing Chapter
  • December 1976

Psychology of Learning and Motivation

... Biological learners also display behaviours that imply the input-independent learning of output statistics. In probability matching, responses mirror the probabilities of rewarded actions [14,15,16]. Learners often display non-stationary biases that are driven by the distribution of recent responses [17,18,19]. ...

Probability Learning11Preparation of this review was supported in part by Contract Nonr 908(16) between the Office of Naval Research and Indiana University. Reproduction in whole or in part is permitted for any purpose of the United States Government.
  • Citing Chapter
  • January 1964

... In addition to identifying conflicting response competition, previous studies indicated that conflict effects in the flanker task were due to perceptual confusion between the target and distractors (Bjork & Estes, 1973;Coles et al., 1985;Estes, 1975). Such confusion has been argued to depend on the similarity between the target and distractors (Moore et al., 2021;Wloka et al., 2017;Yeh & Eriksen, 1984). ...

Letter identification in relation to linguistic context and masking conditions
  • Citing Article
  • September 1973

Memory & Cognition

... In contrast, work that models belief revision as a fractional adjustment process (Bush and Mosteller 1955, Levinthal and March 1981, Denrell and March 2001, whereby the revised beliefs are a mixture of the prior beliefs and whether recent experience is viewed favorably or not, forces an explicit consideration of the rate of learning with respect to beliefs. A partial adjustment process is also more consistent with experimental evidence that shows that more recent events, even in a static problem environment, are weighed more heavily than earlier events (Estes 1972, Hogarth and Einhorn 1992, Camerer and Ho 1999, in contrast to the assumption that beliefs correspond to average experience which weighs all experience equally. Thus, what is often treated as a fairly narrow technical modeling assumption either invites or foreclosures an explicit consideration of a basic mechanism of learning. ...

Research and Theory on the Learning of Probabilities
  • Citing Article
  • March 1972

... Unlike existing methods that attempt to describe isolated voxel activation, favoring models that are correlated with more voxels [18], our novel method favors those models that are predictable by coherent brain states regardless of whether those states are found in activation patterns that reach across the extent of the brain or are localized to specific regions. Of course, this method shares the limitations of any model selection approach, namely that selection is respective to a particular set of tasks, data, and models [28,29]. In the current task and category structure, which has been used as a benchmark for models of categorization [6,[12][13][14], we found evidence strongly in favor of the exemplar model over the prototype model. ...

Human Behavior in Mathematical Perspective
  • Citing Article
  • January 1975

American Scientist

... number of positive trials) of the potential reward (i.e. bait consumed), which influence learning rates Lauer & Estes, 1955;Muzio et al., 1992). Although white sharks can consume baits when operators are unable to detect rapidly approaching sharks, the bait is most often retracted in time, preventing white sharks accessing food cued by scent. ...

Successive acquisitions and extinctions of a jumping habit in relation to schedule of reinforcement
  • Citing Article
  • February 1955

Journal of Comparative and Physiological Psychology

... 3 The foregoing analysis of the current results assumes that generalization will be enhanced, and discrimination impaired, between stimuli (in this case between (a) distractor stimuli and (b) preexposed and nonpreexposed stimuli) to the extent that they share an element, or elements, in common. This is an uncontroversial assertion and one which forms the basis of many models of learning (e.g., Atkinson & Estes, 1963;Hull, 1943;I. P. L. McLaren & Mackintosh, 2002;Rescorla & Wagner, 1972) and which also has been confirmed experimentally (e.g., Kalish, 1958;Mackintosh et al., 1991). ...

Stimulus sampling theory
  • Citing Chapter
  • January 1963

... This is often found where the signal is nil, as in simple probabilistic choice (Laming, 2014;Vulkan, 2000). Sometimes probability matching is treated as a principle, but more interesting treatments derive it as a theorem (Estes, 1957;Thomas, 1975). The prevalence of probability matching has occasioned much research, as the rational choice in these situations is to place all one's responses on the higher probability outcome, as noted in the introduction. ...

Of models and men

American Psychologist

... There is now substantial literature attesting to the accuracy with which absolute and relative frequency judgments are made, both in laboratory (e.g., Hintzman & Block, 1971;Howell , 1973b) and in naturalistic (e.g. , Attneave , 1953;Lichtenstein , Slovic , Fischhoff, Layman , & Combs, 1978) settings. Several theorists have considered ways in which frequencies come to be represented in memory (e.g. , Hintzman, 1976;Howell, 1973a;Whitlow & Estes, 1979); others have discussed heuristics, such as availability, that could underlie frequency judgments (e.g., Combs & Slovic, 1979;Tversky & Kahneman , 1973). ...

Judgments of relative frequency in relation to shifts of event frequencies: Evidence for a limited-capacity model

Journal of Experimental Psychology Human Learning & Memory