1 OF 66

AU Torres-Quevedo-Rocio.

TI A Bayesian approach to estimating a correlation with missing data.

SO Dissertation Abstracts International. 1993 Oct Vol 54(4-B) 2268.



ID maximum likelihood estimates vs Bayesian Highest Density Regions & Bayes Theorem. ************************************************************************

2 OF 66

AU Bornstein-Brian-H.

TI David, Goliath, and Reverend Bayes: Prior beliefs about defendants' status in personal injury cases.

SO Applied Cognitive Psychology. 1994 Jun Vol 8(3) 233-258.



ID prior beliefs & defendant's occupational status, personal injury liability & compensation judgments in simulated jury tasks, college students.

AB Examined the effect of defendants' status (large company vs small business) on judgments by 64 college students in simulated jury tasks. High-status defendants (HSDs) were perceived as less sympathetic and were more likely to lose their case than were low-status defendants. There was a spillover effect such that plaintiffs (Davids) suing HSDs (Goliaths) were themselves viewed as more sympathetic. The effect of defendants' status was mediated by the sentiments it aroused toward both litigants. HSDs did not have to pay more for the same injury. In treating defendants differently depending on status, Ss relied on relevant prior beliefs about the relationship of defendants' status to wealth, propensity to cause harm, and standards of accountability. Results are discussed in terms of the sometimes conflicting norms between an optimal decision-making model like Bayes' Theorem and legal guidelines for judgments of liability and compensation. (PsycINFO Database Copyright 1995 American Psychological Assn, all rights reserved). ************************************************************************

3 OF 66

AU Elwood-Richard-W.

TI Clinical discriminations and neuropsychological tests: An appeal to Bayes' theorem.

SO Clinical Neuropsychologist. 1993 Apr Vol 7(2) 224-233.


ID application of Bayesian probabilities to neuropsychological tests & clinical discriminations.

AB Neuropsychological tests are routinely used to make clinical judgments regarding the classification, diagnosis, or prediction of subject variables. The accuracy of those decisions depends on the discriminant validity of the tests used. The traditional method of validating psychological tests by comparing the mean scores of reference groups is often unrelated to actual clinical discriminations. Alternative measures based on Bayesian probabilities are reviewed, with specific application to neuropsychological assessment. (PsycINFO Database Copyright 1994 American Psychological Assn, all rights reserved). ************************************************************************

4 OF 66

AU Gayol-Luis. Ferreiras-R. Alvarez-J.

TI Estudios familiares en el sindrome fragil X. (Family studies of the fragile X syndrome.).

SO Archivos de Neurobiologia. 1991 Mar-Apr Vol 54(2) 41-47.



ID clinical & cytogenetic & DNA examination, families with sufferers of fragile X syndrome.

AB Using clinical and cytogenetic examination, a study of bearers of the fragile X syndrome was carried out in 5 families with sufferers. DNA analysis techniques were used in the study of 1 of them. The X marker was found in 14 of 20 males; in the 45 females, 23 were heterozygous, of whom 11 were fragile X positive. In the remaining 22, an a priori carrier risk of 25-50% was estimated; however, application of Bayes theorem with conditional probabilities (having a healthy son and being fragile X negative) resulted in a lowering of the bearer risk estimation. DNA study results were concordant with cytogenetic results, and 1 individual manifesting mental retardation, atypical phenotype, and low expression of marker was detected. (English abstract) (PsycINFO Database Copyright 1993 American Psychological Assn, all rights reserved). ************************************************************************

5 OF 66

AU Doherty-Michael-E. Mynatt-Clifford-R.

TI Inattention to P ( H ) and to P ( D H ): A converging operation.

SO Acta Psychologica. 1990 Oct Vol 75(1) 1-11.



ID ability to make probabilistic inferences according to normative principles, college students.

AB Examined people's ability to make probabilistic inferences according to appropriate normative principles (Bayes' theorem). Two investigations were patterned spatially after P. C. Wason's (1968) 4-card task. First, Ss were shown 4 cards, corresponding to the base rate (BR), the complement of the BR, the probability of the datum (POD) given the hypothesis, and the POD given the alternative hypothesis (AHY). In the 2nd study, 3 cards were shown, with the complement of the BR being omitted. Of the 967 university students, only 104 made optimal data choices, and few did so with understanding. These results confirm prior findings (e.g., M. E. Doherty; see PA, Vol 64:4707) that people who are making diagnostic inferences are insensitive to the implications of the BR, and that people are insensitive to the evidentiary impact of the POD given the AHY. (PsycINFO Database Copyright 1991 American Psychological Assn, all rights reserved). ************************************************************************

6 OF 66

AU Ko-Kwangsoo.

TI Information processing of novice and expert in Bayes' Theorem problem-solving.

SO Dissertation Abstracts International. 1990 May Vol 50(11-A) 3656.



ID computational problem solving, novices vs experts in Bayes' Theorem. ************************************************************************

7 OF 66

AU Martin-Scott-L.

TI Honesty testing: Estimating and reducing the false positive rate.

SO Journal of Business & Psychology. 1989 Spr Vol 3(3) 255-267.



ID false positive rate in honesty testing, employee screening tests, application of Bayes' theorem.

AB Bayes's theorem was used to provide a realistic estimate of the false positive rate in honesty testing. Separate estimates were provided for employee theft and production deviance. Examples were used to illustrate how follow-up screening procedures can significantly reduce either the false positive or false negative rate. Suggestions are provided for how test developers can capitalize on research in the area of honesty testing to develop alternative methods of identifying dishonest applicants. (PsycINFO Database Copyright 1990 American Psychological Assn, all rights reserved). ************************************************************************

8 OF 66

AU Kowalczyk-Marek.

TI Prawdopodobienstwo i prototypowosc w procesie kategorialnej identyfikacji percepcyjnej: Analiza teorii spostrzegania Jerome S. Brunera. / Probability and prototypicality in the process of categorical perceptual identification: The analysis of Jerome S. Bruner's theory of perception. (Trans E. Nawrocka).

SO Przeglad Psychologiczny. 1987 Vol 30(2) 319-351.


ID category accessibility & goodness of fit or prototypicality of identified exemplar in categorical perceptual identification.

AB Discusses the interaction between category accessibility (representing the influence of cognitive and motivational extrasensory factors) and the goodness of fit or prototypicality of an identified exemplar in the process of categorical perceptual identification. A mathematical model for processing probabilistic information in the course of category identification is proposed. The implications of the assumption that this model is consistent with Bayes' theorem and E. Rosch's (1978) concept of the structure of natural categories are considered. (English & Russian abstracts) (PsycINFO Database Copyright 1990 American Psychological Assn, all rights reserved). ************************************************************************

9 OF 66

AU Dubois-Didier. Prade-Henri.

TI Modelling uncertainty and inductive inference: A survey of recent non-additive probability systems. 11th Conference on: Subjective probability, utility and decision making (1987, Cambridge, England).

SO Acta Psychologica. 1988 Sep Vol 68(1-3) 53-78.



ID modeling subjective uncertainty & probability theory & inductive inference, implications for artificial intelligence, conference presentation.

AB Modeling uncertainty, especially subjective uncertainty judgments, has been addressed by means of tools from the mathematical theory of probability. However, the discovery of systematic behavioral deviations with regard to the subjective expected utility model, together with the emergence of knowledge engineering, have pointed out some limitations in the knowledge representation power of the standard additive theory. New classes of set functions proposed to model subjective uncertainty are reviewed, and interpretive settings are proposed. The question of preserving notions is discussed with a view to develop new tools for inductive reasoning in the spirit of Bayes theorem. Approaches to inductive reasoning, such as probability kinematics based on information measures and the combination of uncertain or default information as studied in the field of artificial intelligence, are also discussed. (PsycINFO Database Copyright 1989 American Psychological Assn, all rights reserved). ************************************************************************

10 OF 66

AU Frick-Theodore-W.

TI Bayesian adaptation during computer-based tests and computer-guided practice exercises.

SO Journal of Educational Computing Research. 1989 Vol 5(1) 89-114.



ID predictive validity of sequential probability ratio test, assessment of knowledge mastery, graduate students, implications for computer based instruction.

AB Investigated the predictive validity of the sequential probability ratio test (SPRT), developed by A. Wald (1947) and based on Bayes' theorem, for making mastery and nonmastery decisions from 2 pools of test items with varying parameters, using 158 university students. Ss participated in 2 computer-based tests. There was very high agreement between SPRT mastery decisions and decisions reached from total test scores. In the few cases where the SPRT erred, it failed to correctly classify students who were determined to be masters from their total test scores. In no case did the SPRT classify an S as a master who was subsequently determined to be a nonmaster. It is concluded that SPRT is a viable decision model for computer-based instruction when used conservatively with tests suspected or known to contain items of varying difficulty and discriminatory power. (PsycINFO Database Copyright 1989 American Psychological Assn, all rights reserved). ************************************************************************

11 OF 66

AU Einspieler-Christa. Widder-Joachim. Holzer-Andrea. Kenner-Thomas.

TI The predictive value of behavioural risk factors for sudden infant death.

SO Early Human Development. 1988 Dec Vol 18(2-3) 101-109.



ID predictive value of clinical & behavioral risk factors for sudden infant death syndrome, parents of victims.

AB Interviewed the parents of 80 sudden infant death syndrome (SIDS) victims and 80 parents of a healthy control group to examine the predictive value of risk factors for SIDS. These interviews were used to assemble a list of 24 clinical and behavioral symptoms that appeared to be associated with the risk (quantified by Bayes theorem) for SIDS. Out of this list, the average number of symptoms reported to be observable in SIDS victims was twice as large as the average number of symptoms reported for healthy controls. Results show that the following symptoms appeared more often in SIDS victims than in the controls: difficulties in awakening, shrill crying, apathy, few movements during sleep, and cyanosis. It is concluded that behavioral risk factors have a low but still remarkable predictive probability. A behavioral pattern of apathy and sleepiness indicates risk with a high probability. (PsycINFO Database Copyright 1989 American Psychological Assn, all rights reserved). ************************************************************************

12 OF 66

AU Carlson-Richard-A. Dulany-Don-E.

TI Diagnostic reasoning with circumstantial evidence.

SO Cognitive Psychology. 1988 Oct Vol 20(4) 463-492.



ID reasoning process model, decision making & problem solving & belief revision using circumstantial evidence clues, college students.

AB Describes a quantitative model that assumes that Ss revise belief with circumstantial evidence on the basis of a cascaded reasoning process that combines beliefs about 3 premises (the association of a clue and a possible cause, and forward and backward implications of the clue) to revise belief in a causal hypothesis. 36 college students in 3 experiments solved fictional murder mysteries, reporting on each trial a subset of the beliefs specified by the model. Exp 1 demonstrated that the model provides a good account of the reasoning process, over several contexts in which clues are evaluated. Exp 2 provided further evidence on the development of causal models and compared the present model with a model based on Bayes' theorem. Exp 3, a control experiment, demonstrated that the belief assessments used in Exps 1 and 2 did not alter the process of belief revision. The present view is compared with other theories of causal thinking and belief revision. (PsycINFO Database Copyright 1989 American Psychological Assn, all rights reserved). ************************************************************************

13 OF 66

AU Kameda-Tatsuya.

TI / Stereotype-based expectancy and social judgment: Rethinking from a Bayesian perspective.

SO Japanese Journal of Psychology. 1986 Apr Vol 57(1) 27-34.



ID prior stereotypic expectancy, social stereotypes & beliefs, adolescents & adults, Japan.

AB Studied the effect of prior steretypic expectancy on social stereotypes and beliefs, using Bayes' theorem. Human subjects: 202 normal male and female Japanese adolescents and adults. The Ss were asked to infer a target person's attitude toward an atomic-power problem. Half were told in advance that the target person was a member of the Liberal Democratic Party (proexpectancy condition), and half were told that he was a member of the Japanese Socialist Party (conexpectancy condition). Then, the Ss were given a series of the target person's previous relevant utterances that had high or low diagnostic values for inferring his attitude. (English abstract) (PsycINFO Database Copyright 1989 American Psychological Assn, all rights reserved). ************************************************************************

14 OF 66

AU Faigman-David-L. Baglioni-A-J.

TI Bayes' theorem in the trial process: Instructing jurors on the value of statistical evidence.

SO Law & Human Behavior. 1988 Mar Vol 12(1) 1-17.



ID understanding & use of statistical information & Bayesian explanation, subjective probability judgment in hypothetical legal trial, adults acting as jury.

AB Addressed 180 adults' ability to use statistical information as well as their ability to understand and use an expert's Bayesian explanation of that evidence. Ss were presented with a transcript purportedly taken from an actual trial and were asked to make several subjective probability judgments regarding blood-grouping evidence. Results extend to the trial process previous psychological research suggesting that individuals generally underutilize statistical information, as compared to a Bayesian model. Ss in this study generally ignored the expert's Bayesian explanation of the statistical evidence. No relationship was obtained between Ss' understanding of the Bayesian presentation and either verdicts or use of the statistician's conclusions. (PsycINFO Database Copyright 1988 American Psychological Assn, all rights reserved). ************************************************************************

15 OF 66

AU Birnbaum-Michael-H. Mellers-Barbara-A.

TI Bayesian inference: Combining base rates with opinions of sources who vary in credibility.

SO Journal of Personality & Social Psychology. 1983 Oct Vol 45(4) 792-804.



ID information & opinion of other & credibility of opinion, probability judgment, college students.

AB 65 undergraduates made judgments of the probability of an event given base-rate information and the opinion of a source. Base rate and the source's hit and false-alarm rates were manipulated in a within-Ss design. Hit and false-alarm rates were manipulated to produce sources of varied expertise and bias. The base rate, the source's opinion, and the source's expertise and bias all had large systematic effects. Although there was no evidence of a ''base-rate fallacy,'' neither Bayes' theorem nor a subjective Bayesian model that allows for ''conservatism'' due to misperception or response bias could account for the data. Responses were consistent with a scale-adjustment averaging model developed by M. H. Birnbaum and R. S. Stegner (1979). In this model, the source's report corresponds to a scale value that is adjusted according to the source's bias. This adjusted value is weighted as a function of the source's expertise and averaged with the subjective value of the base rate. These results are consistent with a coherent body of experiments in which the same model could account for a variety of tasks involving the combination of information from different sources. (34 ref) (PsycINFO Database Copyright 1984 American Psychological Assn, all rights reserved). ************************************************************************

16 OF 66

AU Kyburg-Henry-E.

TI Rational belief.

SO Behavioral & Brain Sciences. 1983 Jun Vol 6(2) 231-273.



ID connection between philosophical normative investigation of rationality & empirical psychological analysis of belief formation & change.

AB Suggests a threefold connection between the philosophical normative investigation of rationality and the empirical psychological study of belief: (1) Most empirical studies take for granted certain normative constraints, which may or may not be appropriate; if they are inappropriate, they may vitiate the results of the psychological investigation. (2) Philosophical investigations into rationality may provide a useful framework within which psychological investigation can be conducted. (3) If investigators wish to make conclusions about the rationality with which people draw conclusions or apportion their beliefs, they want to be well supported by the evidence. Several issues related to deductive and inductive logic are answered by the author based on a probalistic rule of acceptance and a conception of interval-valued logical probability, according to which probabilities are based on known frequencies. It is suggested that these lead to limited deductive closure, a demand for only limited consistency, and a rejection of Bayes' theorem as universally applicable to changes of belief. 18 commentaries are included and the author's response. (137 ref) (PsycINFO Database Copyright 1984 American Psychological Assn, all rights reserved). ************************************************************************

17 OF 66

AU Beyth-Marom-Ruth. Arkes-Hal-R.

TI Being accurate but not necessarily Bayesian: Comments on Christensen-Szalanski and Beach.

SO Organizational Behavior & Human Performance. 1983 Apr Vol 31(2) 255-257.


ID use of Bayes' theorem vs method of information presentation, probability estimation, criticism of study by J. J. Christensen-Szalanski & L. R. Beach.

AB J. J. Christensen-Szalanski and L. R. Beach (see PA, Vol 58:302) presented Ss with a representative sample of cases, each having or not having 2 characteristics (A and B). Ss estimated the probability of A/B relatively accurately. This was interpreted as indicating that under such conditions, Ss used Bayes' theorem properly. The present authors disagree with this interpretation, and claim that the way the information was presented enabled Ss to make a direct estimate of the relative frequency appropriate for the probability of A/B and the percentage of A in the set of B. Under these circumstances, the question of whether people correctly utilize all the components of Bayes' theorem is completely irrelevant. (5 ref) (PsycINFO Database Copyright 1983 American Psychological Assn, all rights reserved). ************************************************************************

18 OF 66

AU Adelman-Leonard. Donnell-Michael-L. Phelps-Ruth-H. Patterson-John-F.

TI An iterative Bayesian decision aid: Toward improving the user-aid and user-organization interfaces.

SO IEEE Transactions on Systems, Man, & Cybernetics. 1982 Nov-Dec Vol 12(6) 733-743.


ID computer aid based on Bayes' theorem, decision making, military intelligence analysts.

AB Developed an iterative Bayesian decision aid to give users greater control of the decision-making process and thereby increase the likelihood of aid implementation. The aid permits users to modify the initially assessed likelihood ratios and resulting posterior odds for the hypotheses under consideration until they feel comfortable with the inferred implications of the data. The aid was applied to the area of military tactical intelligence analysis and evaluated experimentally by experienced analysts on a training task representative of that decision-making environment. 12 experienced and 18 inexperienced (but trained) analysts working with the aid were (1) better able to discriminate between the most and least likely hypotheses and (2) assigned final likelihoods to the hypotheses that were more similar to normative likelihoods derived from Bayes's theorem than unaided analysts. In addition, analyses of individual analysts revealed large differences in decision-making strategies. These differences demonstrate that users can approach the inference problem using their own decision-making style, thereby ensuring user control over all stages of the decision-making process. (27 ref) (PsycINFO Database Copyright 1983 American Psychological Assn, all rights reserved).

19 OF 66

AU Haisch-Jochen.

TI Die Genauigkeit der Informationsbewertung im Rahmen eines modifizierten ""Attributionswurfels.'' / The accuracy of information evaluation in an extended evaluation cube.

SO Psychologische Beitrage. 1982 Vol 24(1) 26-38.


ID causal attribution task, use of extended judgmental dimensions & applicability of Bayes' theorem as normative decision rule.

AB Two experiments investigated observers' application of extended judgmental dimensions when attributing causes to events and examined whether Bayes' theorem could be applied as a normative decision rule in attribution theory. ''Objective'' diagnostic values of attributional information estimated by experts varied significantly, depending on the information's evidence. Ss performing a causal attribution task failed to evaluate information in such a way as to correspond to the given ''objective'' values. Thus, Bayes' theorem, used as a normative decision rule in attribution theory, seems to involve problems similar to those related to the covariation principle postulated by H. H. Kelley (1967). (French abstract) (28 ref) (PsycINFO Database Copyright 1983 American Psychological Assn, all rights reserved). ************************************************************************

20 OF 66

AU Doscher-Mary-Lynn. Bruno-James-E.

TI Simulation of inner-city standardized testing behavior: Implications for instructional evaluation.

SO American Educational Research Journal. 1981 Win Vol 18(4) 475-489.


ID Bayes theorem procedure for bypassing distortions in test results due to guessing, implications for inner city instructional program planning & evaluation.

AB Examined possible distortions due to guessing in standardized test scores, focusing particularly on inner-city situations. A practical procedure based on Bayes theorem is presented for estimating true knowledge levels from observed test scores. Results show test scores were overstatements of subject mastery, with larger distortions at the lower achievement levels. This suggests that many inner-city programs are being planned and evaluated using data containing large amounts of error. (14 ref) (PsycINFO Database Copyright 1982 American Psychological Assn, all rights reserved). ************************************************************************

21 OF 66

AU Daniel-Wayne-W. Schott-Brian. Atkins-Fred-C. Davis-Alpheus.

TI An adjustment for nonresponse in sample surveys.

SO Educational & Psychological Measurement. 1982 Spr Vol 42(1) 57-67.


ID Bayes theorem, adjustments for nonresponse bias in sample surveys.

AB One of the more serious problems encountered in conducting sample surveys is that of nonresponse. A variety of techniques for dealing with the problem have been proposed, but none have been universally satisfactory. The present study investigates the use of Bayes's theorem in adjusting for nonresponse bias using 600 hospitals in a simulated sample survey. On the basis of known information on 5 variables, Bayes's formula correctly predicted the status of 92 out of 100 ''nonrespondents'' relative to a 6th variable. (43 ref) (PsycINFO Database Copyright 1982 American Psychological Assn, all rights reserved). ************************************************************************

22 OF 66

AU Locksley-Anne. Borgida-Eugene. Brekke-Nancy. Hepburn-Christine.

TI Sex stereotypes and social judgment.

SO Journal of Personality & Social Psychology. 1980 Nov Vol 39(5) 821-831.


ID sex of target & assertive vs passive behavior, prediction of behavior & personality rating along sex stereotypic trait dimensions, college students.

AB Tested the assumption that sexual stereotypic beliefs affect the judgments of individuals in an experiment with 98 male and 97 female undergraduates. No evidence was found for effects of stereotypes on Ss' judgments about a target individual. Instead, Ss judgments were strongly influenced by behavioral information about the target. To explain these results, it is noted that the predicted effects of social stereotypes on judgments conform to Bayes' theorem for the normative use of prior probabilities in judgment tasks, inasmuch as stereotypic beliefs may be regarded as intuitive estimates for the probabilities of traits in social groups. Research in the psychology of prediction has demonstrated that people often neglect prior probabilities when making predictions about people, especially when they have individuating information about the person that is subjectively diagnostic of the criterion. An implication of this research is that a minimal amount of subjectively diagnostic target case information should be sufficient to eradicate effects of stereotypes on judgments. Results of a 2nd experiment with 75 female and 55 male undergraduates support this argument. (24 ref) (PsycINFO Database Copyright 1981 American Psychological Assn, all rights reserved).

23 OF 66

AU Williams-Roger-L. Youssef-Zakhour-I.

TI Race and position assignment in high school, college and professional football.

SO International Journal of Sport Psychology. 1979 Vol 10(4) 252-258.


ID race, position assignment, high school vs college vs professional football players.

AB Attempted to ascertain whether division of labor along racial lines exists at all levels in football--high school, college, and professional--and to devise a statistical technique by which quantitative prediction could be made about assignment of players to particular football positions on the basis of their race. Results indicate that division of labor manifests itself strongly on all 3 levels, with a disproportionately larger number of Blacks assigned to the positions of running back, wide receiver, and defensive halfback. A disproportionately larger number of Whites are being assigned to the positions of quarterback, center, and offensive guard. By means of Bayes' theorem, likelihood ratios are established that significantly enhance the accuracy of predicting position assignment on the basis of race for each sample, from 24% in high school to 93% in professional teams, with the college sample falling in the middle of this range. (Spanish, French, German, & Italian summaries) (5 ref) (PsycINFO Database Copyright 1980 American Psychological Assn, all rights reserved). ************************************************************************

24 OF 66

AU Lad-Frank.

TI Embedding Bayes' theorem in general learning rules: Connections between idealized behaviour and empirical research on learning.

SO British Journal of Mathematical & Statistical Psychology. 1978 Nov Vol 31(2) 113-125.


ID defense of applicability of Bayesian learning procedure, learning rules that permit deviation from idealized behavior models.

AB A response is made to the recent discussions critical of the Bayesian learning procedure on the basis of empirically observed deviations from its prescriptions. Bayes' theorem is embedded in a more general class of learning rules that allow for departure from the demands of idealized rational behavior. Such departures are termed ''learning impediments'' or ''disabilities.'' Some particular forms and interpretations of impediment functions are presented, and consequences of learning disabilities for the likelihood principle, stable estimation, and admissible decision making are explored. Examples of surprising learning behaviors and decision strategies are generated, resulting in deeper understanding of Bayesian learning and its characteristics. (16 ref) (PsycINFO Database Copyright 1980 American Psychological Assn, all rights reserved). ************************************************************************

25 OF 66

AU Lewis-Ann.

TI The early identification of children with learning difficulties.

SO Journal of Learning Disabilities. 1980 Feb Vol 13(2) 102-108.



ID screening method, early identification of children at risk for learning difficulties.

AB Tested a method for the early identification of children at risk for learning difficulties. 161 children participated in a 3-stage study. During Stage 1, Ss were administered the English Picture Vocabulary Test (EPVT); in Stage 2, teachers supplemented information obtained in Stage 1 with the Croydon Checklist; and during Stage 3 (2 yrs after the initial screening), Ss completed a standardized reading test. Several children were misclassified, and the manipulation of cutoffs, using the Bayes theorem, did little to improve the efficiency of the screening procedure. Changing the predictor (e.g., using the EPVT alone) or changing the criterion did not eliminate the misclassifications. It is suggested that individual children were moving from risk to not at risk and vice versa, indicating a need for determining the factors by which change is determined. (27 ref) (PsycINFO Database Copyright 1980 American Psychological Assn, all rights reserved). ************************************************************************

26 OF 66

AU Hesketh-Jose-L.

TI The decision process: A study of four models.

SO Revista Latinoamericana de Psicologia. 1978 Vol 10(1) 37-52.


ID expectancy theory vs lens model vs Bayes theorem vs subjective expected utility model as models of decision making behavior.

AB Briefly discusses decision making in general and presents 4 different models of decision-making behavior: the expectancy theory, the lens model, the Bayes theorem, and the subjective expected utility model. Each model is described and its advantages and shortcomings examined. It is noted that the models share the common assumption that human beings can make accurate probability estimates; their formulations violate H. A. Simon's ''bounded rationality'' principle and therefore their usefulness as decision-making models becomes very limited. (45 ref) (PsycINFO Database Copyright 1980 American Psychological Assn, all rights reserved).

27 OF 66

AU Kvalseth-Tarald-O.

TI Human and Bayesian information processing during probabilistic inference tasks.

SO IEEE Transactions on Systems, Man, & Cybernetics. 1978 Mar Vol 8(3) 224-229.


ID human information processing in Bayes theorem, probabilistic inference task performance.

AB Used basic information theory statistics as quantitative performance measures of human information processing during a probabilistic inference task requiring 3 Ss to estimate the probabilities of a set of events based on sequential data observations. Reanalysis of previously published experimental data revealed that the Ss were able to gain relatively little information about the events from the observed data as compared to the amount of information that the Bayes theorem prescribed as being available. The average subjective uncertainty (conditional entropy) for the set of events was significantly higher than the Bayesian uncertainty, while the data redundancy was lower in the subjective than in the Bayesian case. Additional data observations had only minor impact on the subjective values of these 3 informational measures, especially when compared with the Bayes norm. The human information processing performance improved substantially when a probabilistic information processing system was used. (26 ref) (PsycINFO Database Copyright 1979 American Psychological Assn, all rights reserved). ************************************************************************

28 OF 66

AU Schaefer-Ralf-E.

TI Bayes' theorem as a diagnostic classification tool.

SO Archiv fur Psychologie. 1977 Vol 129(4) 302-318.


ID Bayes' theorem, utility as diagnostic classification tool.

AB Outlines Bayes' theorem as a general purpose statistical classification procedure. In comparison with more traditional statistical approaches, Bayes' theorem is shown to have a number of positive features, e.g., (a) it is not necessary to assume any analytical distribution function of the random variables that figure as input information, (b) the input information may be discrete or continuous or a mixture of both, and (c) information can be processed sequentially. The statistical prerequisite for a legitimate application of Bayes' theorem as a classification aid is discussed and the procedure is applied to a problem. (20 ref) (PsycINFO Database Copyright 1979 American Psychological Assn, all rights reserved). ************************************************************************

29 OF 66

AU Shavelson-Richard-J. Cadwell-Joel. Izu-Tonia.

TI Teachers' sensitivity to the reliability of information in making pedagogical decisions.

SO American Educational Research Journal. 1977 Spr Vol 14(2) 83-97.


ID stories varying in reliability & valence of information, estimates of student's future academic performance & preinstructional & interactive decisions, teachers & nonteachers, application of Bayes' theorem.

AB Models of teaching as a decision-making process have questioned whether teachers make judgments about student performance by following the Bayes theorem or by using simpler heuristic rules that may lead to systematic errors. Although research suggests that people are not Bayesian, Winkler argues that professional decisions may be Bayesian. To determine whether teachers' estimates of student performance are consistent with the Bayes rule, 164 Ss (teachers and nonteachers) were randomly assigned to 1 of 16 cells (2-sup-4 design) and presented with 16 different stories, varying in terms of the reliability and valence of their information about a fictitious student. Ss read the 1st section of the story and were asked to make initial estimates of the student's future academic performance and preinstructional and interactive decisions. They then read the 2nd section and were requested to revise their estimates and decisions based on the additional information. A comparison of the data with a path model representing the Bayes rule showed that Ss may have found information about general ability less relevant to interactive instructional decisions. (22 ref) (PsycINFO Database Copyright 1978 American Psychological Assn, all rights reserved). ************************************************************************

30 OF 66

AU Schaefer-Ralf-E.

TI The evaluation of individual and aggregated subjective probability distributions.

SO Organizational Behavior & Human Performance. 1976 Dec Vol 17(2) 199-210.


ID subjective probability distribution, direct estimation of posterior probabilities vs likelihoods & subsequent aggregation.

AB After extensive training, 11 college students had to assess subjective probability distributions for uncertain quantities. Some of these quantities served as posterior probabilities, others as likelihoods for which, by Bayes' theorem, posterior probabilities could be derived. Two different kinds of evaluation criteria were used: The 1st refers to the goodness of individual assessments, defined as the score of a proper scoring rule. The 2nd criterion refers to the information inherent in the assessments. The question is whether posterior probabilities derived from assessed likelihoods would lead to better results than the direct estimation of posterior probabilities. With respect to the 1st criterion, evaluation of individual assessments, Ss were rather poor in their performance and did not improve over repeated sessions. Furthermore, there were no differences in performance for the different kinds of assessments made. But if the 2nd criterion is used, posterior probabilities derived from assessed likelihoods turned out to be quite good. The direct assessment of posterior probabilities was always inferior. Implications for the training of probability assessors and the construction of information processing aids are discussed. (PsycINFO Database Copyright 1977 American Psychological Assn, all rights reserved). ************************************************************************