Powers of Goodness-of-Fit Tests in Detecting Balanced Mixed Normal Distributions

  title={Powers of Goodness-of-Fit Tests in Detecting Balanced Mixed Normal Distributions},
  author={S. M. Bajgier and Lalit K. Aggarwal},
  journal={Educational and Psychological Measurement},
  pages={253 - 269}
Measurements of a natural, social or psychological phenomenon in many cases come from a sample drawn from a mixture of populations. The presence of a mixed population and its characteristics however may be a-priori unknown. This state of ignorance can lead to bias in a summary measure of a phenomenon. In this paper, a test based on sample kurtosis is shown to be more powerful than other known tests in detecting a class of mixed normal distributions. 
On the meaning and use of kurtosis.
For symmetric unimodal distributions, positive kurtosis indicates heavy tails and peakedness relative to the normal distribution, whereas negative kurtosis indicates light tails and flatness. ManyExpand
Variable Screening for Cluster Analysis.
Inclusion of irrelevant variables in a cluster analysis adversely affects subgroup recovery. This paper examines using moment-based statistics to screen variables; only variables which pass theExpand
Univariate Screening Measures for Cluster Analysis.
  • J. Donoghue
  • Mathematics, Medicine
  • Multivariate behavioral research
  • 1995
Using moment-based statistics to screen variables; only variables which pass the screening are then used in clustering, and screening variables appears a viable alternative to both ultrametric weights and forward selection. Expand
Two families of kurtosis measures
Abstract. Two families of kurtosis measures are defined as K1(b)=E[ab−|z|] and K2(b)=E[a(1−|z|b)] where z denotes the standardized variable and a is a normalizing constant chosen such that theExpand
An analysis of quantile measures of kurtosis: center and tails
The consequences of substituting the denominator Q3(p)  −  Q1(p) by Q2  −  Q1(p) in Groeneveld’s class of quantile measures of kurtosis (γ2(p)) for symmetric distributions, are explored using theExpand
Practical Semiparametric Inference With Bayesian Nonparametric Ensembles
This thesis proposes Bayesian Nonparametric Ensemble (BNE), a general modeling approach that combines the a priori information encoded in candidate models using ensemble methods, and then addresses the systematic bias in the candidate models use Bayesian nonparametric machinery. Expand
K-means clustering: a half-century synthesis.
  • D. Steinley
  • Computer Science, Medicine
  • The British journal of mathematical and statistical psychology
  • 2006
This paper synthesizes the results, methodology, and research conducted concerning the K-means clustering method over the last fifty years, leading to a unifying treatment of K-Means and some of its extensions. Expand
The use of cluster analysis for cell typing.
This article describes a statistical approach known as cluster analysis that was used to define cell types in the nucleus of the solitary tract-a nucleus that had been suggested to contain identifiable cell types but within which cell typing had proven difficult. Expand
A multivariate approach to structural heterogeneity of retinal ganglion cells
The formal conclusion is that functionally distinct frog AOS GCs do not differ in morphology or dendritic arbor orientation. Expand
Medical representatives bear the blunt of the significant pressure from the peers and managers to accomplish their market targets which call for an aggressive promotion. This inevitably leads toExpand


A Test of Goodness of Fit
Some (large sample) significance points are tabulated for a distribution-free test of goodness of fit which was introduced earlier by the authors. The test, which uses the actual observations withoutExpand
A Simulation Analysis of the Power of Several Tests for Detecting Heavy-Tailed Distributions
Abstract This article presents estimates of the power of five tests for detecting heavy-tailed distributions with the null hypothesis of normality and alternative hypotheses of each of eleven membersExpand
A Monte Carlo Investigation Of The Likelihood Ratio Test For The Number Of Components In A Mixture Of Normal Distributions.
  • B. Everitt
  • Mathematics, Medicine
  • Multivariate behavioral research
  • 1981
The results show that the proposed sampling distribution of the test appears to be appropriate only for sample sizes above fifty, and for data where the sample size is ten times the number of variables. Expand
Tests for departure from normality: Comparison of powers
SUMMARY In the present paper we present results of a large scale simulation study of the power of various tests of normality. We follow the procedure adopted by Shapiro, Wilk & Chen of applying aExpand
The Power of the Shapiro-Wilk W Test for Normality in Samples from Contaminated Normal Distributions
Abstract The power of the W statistic for testing normality proposed by Shapiro and Wilk [3] was studied by means of a sampling experiment for samples drawn from contaminated normal distributionsExpand
An empirical investigation of the properties of a new goodness-of-fit test
We investigate the power properties of a new goodness-nf-fit test proposed by rontz (1980). This new test is compared with the (The squared test and the Kolmogoma Smirnov (K-S) test for normalityExpand
Testing normality of errors in regression models
It was shown by Pierce & Kopecky (1979) that tests of normality appropriate for the identically distributed case are asymptotically valid when computed from residuals in regression situations.Expand
Some Methods for Studying the Mixture of Two Normal (Lognormal) Distributions
Abstract This article explores several topics related to the mixture of two normal or lognormal distributions. A study is made of diagnostic plotting procedures for the detection of mixtures, and aExpand
An Approximate Analysis of Variance Test for Normality
Abstract This article presents a modification of the Shapiro-Wilk W statistic for testing normality which can be used with large samples. Shapiro and Wilk gave coefficients and percentage points forExpand
An Analysis of Variance Test for Normality (Complete Samples)
The main intent of this paper is to introduce a new statistical procedure for testing a complete sample for normality. The test statistic is obtained by dividing the square of an appropriate linearExpand