摘要:

An historical and theoretical review is provided of so called item response theory (IRT), more accurately described as item response modelling (IRM). This paper looks at 50 years of IRM and finds a disappointing lack of advance. It is shown how a linear model framework, involving different response transformations, unifies separate approaches to the study of test item responses.

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00905.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

In investigations into the manner in which people store spatial information, experimental subjects are often asked to estimate the distances and bearings between various landmarks. Methodology is described for constructing the mental map underlying a subject's matrix of distances, and the mental map underlying a subject's matrix of bearings. Also presented are methods of comparing matrices of distances or bearings, and the mental maps constructed from them, with the geographical map; this enables an assessment of differences between them. Some statistical models of disorientation are considered, and a method of choosing between them is described.

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00906.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

Simulations were performed to compare the power of the approximate permutation test with the power ofttest and Wilcoxon's test for the two‐sample location problem under a shift model. The approximate permutation test is sometimes suggested as a panacea for non‐normality. However, for the distributions and sample sizes used in this study, the power of the approximate permutation test and thettest are nearly equal. Under non‐normality Wilcoxon does have better power characteristics than the other tests. So it can be concluded, that in this study Wilcoxon, a permutation test on ranks, does perform better under non‐normality than the approximate permutation test that uses the measurements the

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00907.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

Observational studies in applied psychology raise particular problems in reliability measurement. Existing techniques for measuring agreement on nominal scales have concentrated on the case where each rater rates all the ratees. The less common case where each ratee has been judged by a different pair of raters has been discussed by Fleiss (1971, 1981) as part of the more general problem where multiple different raters rate each case. This note shows how the formula may be re‐expressed for two different raters per case in a way which leads to the use of existing software for the usualK.This identity may also be useful didacticall

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00908.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

Goodman&Kruskal (1954) introduced a measure λ of predictive association when predicting the category of a variableAfrom a category of a variableB.The measure λ is such that 0≤λ≤1 with a zero value meaning no predictive gain, while λ=1 indicates a perfect predictive association betweenAandB.Turek&Suich (1983) developed an exact test ofH0:λ = 0 versusH1:λ>0where the variableAis dichotomous. This test is analogous to a test for the significance of the correlation coefficient. This paper analyses the partial λ coefficient in order to answer the question of whether knowledge of a third (fourth, fifth, etc.) classification results in a significant increase in ability to predict the dichotomous variableA.This test, developed for qualitative categorical variables, is analogous to the test for partial correlation coefficients for quantitative

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00909.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

Research has shown that when an appropriate data transformation is knowna priori, then it can lead to a substantial increase in power of analysis of variance while maintaining an appropriate Type I error rate. It was unknown, however, whether data transformation selected on sample characteristics would yield accurate Type I error rates and increased power. The present Monte Carlo study demonstrates that correct data transformation values could be selected on samples as small as four per group, that legitimate approaches do not inflate the nominal significance levels and that power could be increased by sample‐based transformation

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00910.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

A general two‐level model for multivariate data is described and illustrated by specializations to a common factor model and a path model with latent variables. The problem of estimation is treated, and in the special case of a balanced sampling design, a likelihood‐based discrepancy function and a test of goodness of fit are obtained in terms of some simple sufficient statist

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00911.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

Previous results presented in Tanaka&Huba (1985) introduced a general Fit index for covariance structure models under generalized least squares (GLS) estimation which, in some cases, specialized to the fit indices presented in Jöreskog&Sörbom (1981). Here, it is shown that, for a wide class of models, the general form of this fit index can be expressed as a weighted coefficient of determination. This coefficient is given as the ratio of weighted trace functions of predicted and observed covariance matrix element

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00912.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

A new approach is proposed for data which are skewed and have a sizeable proportion of observation at variable end points. Using a covariance structure modelling framework, the new approach assumes censored multivariate normal variables. Using bivariate information, this leads to the use of ‘tobit’ correlations in weighted least squares estimation. The behaviour of the tobit approach is compared to that of normal theory estimation and ADF estimat

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00913.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY

摘要:

When a multiple choice test is carried out, each student is set the same series of questions composed of two or more alternatives, only one of which is correct, and from which he or she must choose only one. The simplest and most common criterion for deciding what a student knows is based on a sole and minimum quota for all the students of the total number of right answers; this criterion derives from the assumption that the student, when he or she does not know the correct answer to a question, answers any of the possible alternatives with equal probability.In this paper, the authors show that this method may be unsatisfactory if the student is unable to imitate random selection, and they put forward an alternative model which allows one to find an estimator for the student's level of knowledge that depends on the sum of the proportions of correct answers of each type and on the total number of each type which the student gives. Likewise, they give a hypothesis test for verifying the model, two methods for finding confidence intervals for the student's level of knowledge, and three hypothesis tests for awarding a mark for it.Numerical calculations show how inefficient the usual multiple choice tests are (100 or 150 questions with five alternative answers each) for ascertaining the student's level of knowledge.

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1989.tb00914.x

出版商:Blackwell Publishing Ltd    

年代:1989

数据来源: WILEY