摘要:

The paper contains a proof that the Guttman additive quasi‐simplex model is invariant under selection of examinees when a set of continuous variables each having different error of measurement is ordered on the simplex continuum in terms of either increasing or decreasing magnitudes of ‘complexity’ provided that selection is based on those measures which are less variable than each of the measures of the remaining set. When variables with highest ‘complexity’ ranks constitute the basis of selection, then such an invariance breaks down. The significance of the problem is discussed in the context of personnel selection. In addition, many implications of the results are stated which hold good for various multivariate statistical t

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00417.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

An approximate method, with decision‐based error‐rates, for testing linear statistical hypotheses (0κi) aboutaindependent binomial parameters (φj) is given which is more general than the usual methods for partitioningXm2from a 2 ×atable. Parameters in Table 2 may be used to choose sample‐size (n) for stated alternatives to 0κiand power (or expected proportion of rejections). The procedure may be used for planned tests (v1 = 1) orpost‐hoctests (v1 =a− 1 orv1 =a) and in the latter case the probability of rejectingrof the 0κi(r= 1, 2, …,v1) can be directly calculated for any statement of what is true in the populations.The results from sampling experiments are used to examine the adequacy of the approximation and compare the method proposed (Rv1) with two others (Gv1 andZv1) Although the latter methods have a power advantage under some circumstances, their experimental results forn≤ 30 are generally less close to ‘theory’ than those ofRv1, and they are always arithmetically and conceptually more complex. The results from these experiments suggest that theRv1 approximation is adequate whenn≥ 10 and the row of smaller expected frequencies (nφ) is greater than or equal to 1. When the φjdiffer and the true φjapproach 0 or 1, “end‐effects” appear but these diminish asnincreases.These methods can be used in a 2×ageneralization of the 2 × 2 median test. If this is interpreted not as a test of population medians, but as a test of proportions of population distributions to the right ofMe(the value of the common sample median), which is φjin

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00418.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

Formulae are developed from the assumptions of classical test theory to demonstrate the effect of error of measurement on the power of theFtest for a fixed‐effects one‐way analysis of variance model. A simple model specifying the fixed and variable costs of testing is adopted and equations are derived that indicate the sample size that maximizes the non‐centrality parameter subject to the cost constraints. Given the sample size, the corresponding test length is then implied by the cost model. A computer program used for the estimation of power for permissible allocations of resources is desc

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00419.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

A statistical design and model is presented for estimating the after‐effects of the treatment given at trialnon trialn+ 1, when there are two kinds of treatment in a series. The specific application of the model is to an evaluation of the effects of stimulation produced by reward and non‐reward in a partial reinforcement situation. A rationale for the model is provided; the derivation of the estimates and the analysis of variance is described; and the model is applied to the analysis of some data which form part of a study published elsewhere. The results of the analysis indicate that reward and non‐reward affect immediately following performance differentially at 20 sec. intertrial intervals but not at 12 min. intertrial inte

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00420.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

摘要:

The nature of aggression or hostility is discussed, and the possible relations between hostility in the personalities of husbands and hostility in the personalities of wives are conceptualized under the heads: (a) Similarity in quantity of aggression; (b) Similarity in pattern of aggression; (c) Oppositeness or complementarity in quantity of aggression; (d) Oppositeness or complementarity in pattern of aggression; and (e) Systematic relations between similar or dissimilar (but not necessarily opposite) patterns of aggression. The first four possibilities are explored by imposing a model on questionnaire data obtained from alcoholics and their wives. The fifth possibility, which is the first to be dealt with, is investigated with the aid of a technique which follows the pattern of the data without imposing a model on them.The analysis demonstrates that exclusive reliance on correlational methods, however complex, may obscure the essential features of a psychological situation. It also shows how the special characteristics of temperamental qualities may be accommodated by the construction of an appropriate model.

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00421.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00422.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY

ISSN:0007-1102    

DOI:10.1111/j.2044-8317.1969.tb00423.x

出版商:Blackwell Publishing Ltd    

年代:1969

数据来源: WILEY