摘要:

The knowledge state of an individual with respect to a particular body of information is conceptualized as the set of all the questions that this individual is capable of solving. The goal of an assessment procedure is to identify, by a sequence of appropriately chosen questions, the individual's state among all possible ones. A deterministic procedure is conceivable, but not realistic, in that it does not account for possible inconsistencies in the observed responses. Such inconsistencies may arise from careless errors or lucky guesses from the subject, but may also be of a more fundamental character. A stochastic framework is developed here, in which an individual ‘state’ is formalized as a distribution on the set of all possible knowledge states. On each trial, the assessor has a likelihood function on the set of knowledge states which provides the basis for selecting the question to be asked on that trial. The response is assumed to depend on the individual ‘state’. The question asked and the response observed are used to update the likelihood function. Several examplesof questioning rulesandof updating rulesare discussed, which lead to Markov processes. A central problem is to describe conditions ensuring that the latent distribution corresponding to the subject's ‘state’ can be estimated. We show that, under general conditions, the ‘state’ of a subject is uncoverable if the latent distribution is concentrated on a particular

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00884.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

摘要:

This paper deals with a theory for two variants of the method of triads—Torgerson's and Richardson's methods. For both methods, the relationship between the decision probabilities and the parameters of the momentary psychological magnitudes is derived under the assumption that these magnitudes can be modelled as if they were drawn from normal distributions with particular means and variances. This approach differs from previous theoretical work on the methods of triads, where it has been assumed that distances between momentary psychological magnitudes are normally distributed. Cases where the variances of the distributions of psychological magnitudes are unequal can be handled by the models. It is shown that the duo‐trio and triangular methods are special cases of Torgerson's and Richardson's methods of triads, respectively. Using non‐linear least squares minimization, estimates of the means of the distributions of a sample problem are obtained for random samples of size 200. Decision conflicts, which may occur when using Richardson's method, are disc

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00885.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

摘要:

The problem of selecting a subset of stimuli when empirical confusion frequencies are given is discussed for a case in which a quartet of letter pairs must be found with maximum similarity within pairs and minimum similarity between pairs. As a preliminary step the Shepard–Luce model is fitted; a non‐metric three‐dimensional scaling analysis provides confirmation of the validity of the actually selected

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00886.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

摘要:

This paper considers different ways of estimating variance components, and hence of estimating generalizability coefficients, with particular application to systematic observation studies which use time sampling. Recent work on multilevel models is reviewed and these models are shown to have advantages over other methods for data of this type. The ideas are applied to a recent study of children observed in infant school classes.

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00887.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

摘要:

A simple iterative algorithm is proposed for obtaining the maximum‐likelihood estimates (m.l.e.) of the parameters of the symmetric Toeplitz or Laurent correlation matrix. A numerical example based on real‐life data is provided to illustrate the computation involved in the estimation and to demonstrate the feasibility of the proposed algorithm. This paper also provides expression in explicit form for the computation of the asymptotic dispersion matrix of the MLE. For comparative purposes, the Toeplitz correlation hypothesis as also other structural hypotheses have been tested and the results are discussed. The use of the proposed method in estimating the elements of Σ in the first step and then the possibility of using this estimated Σ for deriving a generalized least square estimate of the regression coefficients is pointe

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00888.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

摘要:

Random effects latent trait models involve an arbitrary assumption about the form of the prior distribution of the latent variable(s). Although the form of the analysis does not depend on this assumption, it does affect the estimates of the parameters. It is therefore important to know whether the values of the estimates obtained are sensitive to the choice made. This question is answered mainly by providing numerical evidence that the effect of the prior on the expected one‐ and two‐way marginal proportions is negligi

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00889.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

摘要:

A new procedure is proposed for testing the hypothesis that the means ofJindependent normal distributions are equal. Extensive simulations indicate that the new procedure always gives the experimenter more control over the probability of a Type I error than does the Brown–Forsythe or Welch adjusted degree of freedom techniques. For equal and reasonably large sample sizes the Welch procedure might be preferred, however, because it can have substantially more power, and it provides adequate though liberal control over Type I errors. However, Welch's procedure cannot always be recommended for reasons described in the paper. For unequal sample sizes both the Welch and Brown–Forsythe procedures are known to be unsatisfactory in certain practical situations, while the new procedure has Type I error probabilities that are much closer to the nominal level. Moreover, the new procedure seems to give conservative results provided the sample sizes in each group are greater than or equal to 10. New results are also reported on James's second order method. Although computationally more tedious, James's procedure is recommended for general

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00890.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

摘要:

A model for the adaptive formation of an associative memory structure based on a stochastic approximation algorithm is described. This model incorporates a feedback loop which allows the modification to memory to depend upon the novelty of the stimulus input. Predictions are derived for change in the memory structure across trials and these results are applied to the analysis of serial position effects in the Sternberg item‐recognition paradigm as well as familiarity judgements in a continuous recognition task. Illustrative applications of the model arc presente

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00891.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

摘要:

Book reviewed in this article:Directory of Statistical Microcomputer Software, 1988 Edition.Compiled by W. A. Woodward, A. C. Elliot, H. L. Gray&D. C. Matlock.Natural and Artificial Intelligence. Processor Systems Compared to the Human Brain.ByArmand M. de Callataÿ.The Cult of Information. The Folklore of Computers and the True Art of Thinking.ByTheodore Roszak

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00892.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1988.tb00893.x

出版商:Blackwell Publishing Ltd    

年代:1988

数据来源: WILEY