摘要:

In recognition tasks where the stimuli vary along two or more dimensions, the question arises as to what strategy the subject uses in processing the information about the various stimulus attributes. Three strategies are discussed: parallel, serial and integrative processing. It is shown that each of these three strategies implies a relation among the distribution functions of three observable random variables. The problem is then to decide on the basis of three samples, one from each distribution, which one of the relations is most likely. Non‐parametric statistics are defined which can be used to construct test functions for this decision problem, and a suggestion is made concerning the design of experiments which attempt to discriminate between these strategie

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00424.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

An experiment was carried out to investigate the effect of the number of response alternatives on response latency in paired‐associate learning. In previous studies no systematic differences have been found between error and success latencies of trials prior to the last error, but these studies have involved only two or three response alternatives. In the present study eight stimuli were paired with either two‐, four‐ or eight‐response alternatives.The mean error latencies were an increasing function of the number of response alternatives, while on pre‐criterion trials, mean error latencies were consistently greater than mean success latencies with four‐ and eight‐response alternatives. No systematic differences were found in the two‐response condition.A four‐state all‐or‐none memory model was fitted to both the frequency and latency data, and adequately accounted for the four‐ and eight‐response Conditions, but yielded some discrepant predictions in the two‐response condition. These discrepancies were attributed to the use of different recall

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00425.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

The relative efficiencies of various experimental designs are compared for the genetical analysis of human behaviour. Particular attention is paid to the ability of the different designs to separate additive genetic variation from dominance variation. The utility of twins is investigated and it is shown that the relative number of twins need not be large for the statistical efficiency of the experiment to be optimal. The possibility of using foster‐siblings in the place of monozygotic twins reared apart is considered. It is demonstrated that a basic and efficient genetical analysis may be undertaken without recourse to monozygotic twins reared apart. The consequences for efficiency of limited availability of certain groups are investigated, and the method of the study illustrated with the aid of a worked exampl

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00426.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

An algorithm is described, the purpose of which is to express a correlation matrix as the sum of two matrices, of which one is of rank less than its order, and the other, the residual matrix, is of prescribed structure. Applications are described to approximate simplex and circumplex matrices, to the factoring of groups of variables, and to a form of multi‐mode factor analysi

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00427.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

An algorithm is described that serves to fit a linear latent structure model to multicategory data. The model has the formal properties of a (generalized) common factor model, in contrast to principal component analysis. An empirical example is given.

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00428.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

Two forms of ac‐sample normal scores test are presented as alternatives to the Kruskal–Wallis test (KW). The normal scores test statistics and multiple comparison procedures are illustrated for large samples requiring a correction for tied observations. Theoretical and empirical evidence is cited in support of using the normal scores test in preference to a rank test such as the KW. The normal scores test is asymptotically more efficient than the KW for samples from common non‐normal populations (e.g. uniform and exponential) and is less sensitive to non‐normality and unequal variances than is either theF‐test or th

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00429.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

A simple relation between error rates using multiplet‐tests and Scheffé's multiple‐comparison procedure is pointed out. Errors in the development of some new multiple‐comparison procedures are indicated. An experimental error rate approach using a flexible error rate is adv

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00430.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

A computer system is described which carries out on‐line experiments involvingm‐independent variables and one dependent variable. Using a search procedure that applies hillclimbing techniques, the system selectively places points in regions that demand them for reliable description. Some results of its performance are given in Section 3. The experimentation system leaves information in store and this is used by a model‐assessment procedure in order to achieve the optimum predictive efficiency for each of the functional models under consideration.This study is part of a research project aimed at producing computer‐controlled experimentation which will not only assess functional models, but also generate them, assess and generate discrete or ‘structural’ models, and design and carry out experiments whose results will be essential for discrimination between models. The work is also intended as a contribution to the problem of computer progra

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00431.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY