Robustness of Multiple Comparison Procedures: Treatment versus Control

Computer simulation techniques were used to investigate the Type I and Type II error rates of one parametric (Dunnett) and two nonparametric multiple comparison procedures for comparing treatments with a control under nonnormality and variance homogeneity. It was found that Dunnett's procedure is quite robust with respect to violations of the normality assumption. Power comparisons show that for small sample sizes Dunnett's procedure is superior to the nonparametric procedures also in non-normal cases, but for larger sample sizes the multiple analogue to Wilcoxon and Kruskal-Wallis rank statistics are superior to Dunnett's procedure in all considered nonnormal cases. Further investigations under nonnormality and variance heterogeneity show robustness properties with respect to the risks of first kind and power comparisons yield similar results as in the equal variance case.     

Testing the robustness of the likelihood-ratio test in a variance-component quantitative-trait loci-mapping procedure.

Detection of linkage to genes for quantitative traits remains a challenging task. Recently, variance components (VC) techniques have emerged as among the more powerful of available methods. As often implemented, such techniques require assumptions about the phenotypic distribution. Usually, multivariate normality is assumed. However, several factors may lead to markedly nonnormal phenotypic data, including (a) the presence of a major gene (not necessarily linked to the markers under study), (b) some types of gene x environment interaction, (c) use of a dichotomous phenotype (i.e., affected vs. unaffected), (d) nonnormality of the population within-genotype (residual) distribution, and (e) selective (extreme) sampling. Using simulation, we have investigated, for sib-pair studies, the robustness of the likelihood-ratio test for a VC quantitative-trait locus-detection procedure to violations of normality that are due to these factors. Results showed (a) that some types of nonnormality, such as leptokurtosis, produced type I error rates in excess of the nominal, or alpha, levels whereas others did not; and (b) that the degree of type I error-rate inflation appears to be directly related to the residual sibling correlation. Potential solutions to this problem are discussed. Investigators contemplating use of this VC procedure are encouraged to provide evidence that their trait data are normally distributed, to employ a procedure that allows for nonnormal data, or to consider implementation of permutation tests.     

A Monte Carlo Test for Variance Homogeneity in Linear Models

A Monte Carlo procedure is proposed for testing homogeneity of variances in linear models. The method is applicable to a variety of common experimental designs. It is valid when errors are independently normally distributed. Under nonnormality the test is expected to behave robust in a similar fashion as Levene's test. Three examples are given to demonstrate the method.     

Rapid Diagnosis of Infection by Gas-Liquid Chromatography: Analysis of Sugars in Normal and Infected Cerebrospinal Fluid

A highly reproducible procedure was developed for gas-liquid chromatographic analysis of trimethylsilyl derivatives of normal human cerebrospinal fluid. Fourteen normal human cerebrospinal fluid samples tested with this procedure contained alpha- and beta-glucose as well as isomers of two unidentified sugars. Chromatographic changes in three cases of meningeal inflammation (two cryptococcosis and one thalamic astrocytoma) were limited to decreased concentrations of all sugars. In one case of early meningitis, the concentrations of the unknown sugars decreased before glucose. Now that a reproducible chromatogram of the trimethylsilyl derivatives of normal human cerebrospinal fluid has been established, more samples of abnormal cerebrospinal fluid should be prepared by these methods and examined by gas-liquid chromatography. It may be possible to identify unique products of infectious agents which will permit rapid diagnosis of central nervous system infection.     

Testing the robustness of the new Haseman-Elston quantitative-trait loci-mapping procedure

Variance components (VC) techniques have emerged as among the more powerful methods for detection of quantitative-trait loci (QTL) in linkage analysis. Allison et al. found that, with particularly marked leptokurtosis in the phenotypic distribution and moderate-to-high residual sibling correlation, maximum likelihood (ML) VC methods may produce a severe excess of type I errors. The new Haseman-Elston (NHE) method is a least-squares-based VC method for mapping of QTL in sib pairs (Elston et al.). Using simulation, we investigate the robustness of the NHE to marked nonnormality, by means of the same distributions and worst-case conditions identified by Allison et al. for the ML approach (i.e., 100 pairs; high residual sibling correlation). Results showed that, when marked nonnormality is present, the NHE can be used without severe type I error-rate inflation, even at very small alpha levels.     

Commingling in the distributions of fat-related measures in Punjabi families

The distribution of anthropometric measurements related to fatness levels is examined to determine if skewness alone accounts for the nonnormality of such measures. A mixture of two normal distributions or a single skewed distribution fit the data significantly better than a single normal in all cases. For maximum hip width, knee diameter, and weight, two skewed distributions give a better fit than one skewed distribution, rejecting the null hypothesis of a single distribution even when skewness is considered. There is evidence for three skewed component distributions for biceps skinfold. Abdomen circumference, upper arm circumference, triceps skinfold, and calf skinfold are best approximated by a one component log-normal distribution. Children and parents show slightly different patterns in skewness and kurtosis when considered separately, but differences between them do not account for the commingling found in the combined distributions.     

正态总体均值与标准差比在半序约束下的最大似然估计

考虑两个正态总体X_i～N(μ_i,σ_i~2) i=1,2,均值μ_i和方差σ_i~2>0是未知参数,它们的比在半序约束μ_1/σ_1≤μ_2/σ_2下的最大似然估计(MLE)问题,给出了计算MLE的方法,对三个正态总体的情况,给出了相关的结果,这些方法在生物遗传和药物检验等方面都有广泛应用。     

The Construction of Normal Expectations

The gradual upward changes of standards in normal everyday life have significant environmental implications, and it is therefore important to study how these changes come about. The intention of the article is to analyze the social construction of normal expectations through a case study. The case concerns the present boom in bathroom renovations in Denmark, which offers an excellent opportunity to study the interplay between a wide variety of consumption drivers and social changes pointing toward long-term changes of normal expectations regarding bathroom standards. The study is problem-oriented and transdisciplinary and draws on a wide range of sociological, anthropological, and economic theories. The empirical basis comprises a combination of statistics, a review of magazine and media coverage, visits to exhibitions, and qualitative interviews. A variety of consumption drivers are identified. Among the drivers are the increasing importance of the home as a core identity project and a symbol of the unity of the family, the opportunities for creative work, the convenience of more grooming capacity during the busy family's rush hours, the perceived need for retreat and indulgence in a hectic everyday life, and the increased focus on body care and fitness. The contours of the emerging normal expectations are outlined and discussed in an environmental perspective.     

Expected Error Rates for Logistic Regression Versus Normal Discriminant Analysis

The expected error rates associated with using the allocation rule based on logistic regression are derived in the context of two multivariate normal populations with a common covariance matrix and compared with the corresponding error rates of the classical rule based on this normality assumption. It is shown in terms of the actual sizes of the asymptotic expected error rates that the performance of the logistic procedure does not fall far short of the normality based method, even for widely separated populations. This latter result is not obvious from previously available work on the asymptotic relative efficiency of the logistic procedure.     

Properties of Normal Phylogenetic Networks

A phylogenetic network is a rooted acyclic digraph with vertices corresponding to taxa. Let X denote a set of vertices containing the root, the leaves, and all vertices of outdegree 1. Regard X as the set of vertices on which measurements such as DNA can be made. A vertex is called normal if it has one parent, and hybrid if it has more than one parent. The network is called normal if it has no redundant arcs and also from every vertex there is a directed path to a member of X such that all vertices after the first are normal. This paper studies properties of normal networks. Under a simple model of inheritance that allows homoplasies only at hybrid vertices, there is essentially unique determination of the genomes at all vertices by the genomes at members of X if and only if the network is normal. This model is a limiting case of more standard models of inheritance when the substitution rate is sufficiently low. Various mathematical properties of normal networks are described. These properties include that the number of vertices grows at most quadratically with the number of leaves and that the number of hybrid vertices grows at most linearly with the number of leaves.     

Discretizing a Normal Prior for Change Point Estimation in Switching Regressions

Bayes decision procedures are considered for change point estimation in the simple bilinear segmented model. A discretized normal prior density is employed as the prior distribution for the change point index. Posterior probability functions are developed for this index under a vague prior formulation on the regression parameters. The procedure is applied to an example involving mercury toxicity data.     

Staining Normal and Experimental Motor Nerve Terminals with Tetrazolium Salts

Nitroblue tetrazolium (NBT) has been used to stain motor nerve terminals and unmyelinated axons in vertebrate skeletal muscle, but undesirable background connective tissue coloration resulted. This procedure was improved by separation of the tetrazolium salt's binding from its subsequent reduction. By uncoupling the binding and reduction steps it was possible (1) to improve nerve terminal staining by using tetranitroblue tetrazolium (TNBT), (2) to counterstain and postfix in osmium tetroxide and (3) to enhance the overall tissue preservation. The separate binding and reduction procedure is compatible with postsynaptic acetylcholinesterase staining. Experimentally manipulated and diseased preparations can be successfully stained, and the requirements for optimal staining in each case are described.     

Serum Iron and Iron-Binding Capacity in Normal Subjects

Normal values for serum iron and unsaturated iron-binding capacity of serum were established in a carefully selected group of control subjects by a modification of the procedure described by Schade and coworkers. No significant difference was found between mean values in males and females, and there was no correlation between serum iron levels and age in adults. The value of serum iron as a diagnostic aid is limited by the wide range in normal subjects and the large variation that occurs in the same individual from hour to hour and day to day.     

Robustness to failure of assumptions of tests for a common slope amongst several allometric lines--a simulation study

In allometry, researchers are commonly interested in estimating the slope of the major axis or standardized major axis (methods of bivariate line fitting related to principal components analysis). This study considers the robustness of two tests for a common slope amongst several axes. It is of particular interest to measure the robustness of these tests to slight violations of assumptions that may not be readily detected in sample datasets. Type I error is estimated in simulations of data generated with varying levels of nonnormality, heteroscedasticity and nonlinearity. The assumption failures introduced in simulations were difficult to detect in a moderately sized dataset, with an expert panel only able to correct detect assumption violations 34-45% of the time. While the common slope tests were robust to nonnormal and heteroscedastic errors from the line, Type I error was inflated if the two variables were related in a slightly nonlinear fashion. Similar results were also observed for the linear regression case. The common slope tests were more liberal when the simulated data had greater nonlinearity, and this effect was more evident when the underlying distribution had longer tails than the normal. This result raises concerns for common slopes testing, as slight nonlinearities such as those in simulations are often undetectable in moderately sized datasets. Consequently, practitioners should take care in checking for nonlinearity and interpreting the results of a test for common slope. This work has implications for the robustness of inference in linear models in general.     

Spectrum of Cytogenomic Abnormalities Revealed by Array Comparative Genomic Hybridization on Products of Conception Culture Failure and Normal Karyotype Samples

Approximately 30% of pregnancies after implantation end up in spontaneous abortions, and 50% of them are caused by chromosomal abnormalities. However, the spectrum of genomic copy number variants(CNVs) in products of conception(POC) and the underlying genedosage-sensitive mechanisms causing spontaneous abortions remain largely unknown. In this study, array comparative genomic hybridization(a CGH) analysis was performed as a salvage procedure for 128 POC culture failure(POC-CF) samples and as a supplemental procedure for106 POC normal karyotype(POC-NK) samples. Chromosomal abnormalities were detected in 10% of POC-CF and pathogenic CNVs were detected in 3.9% of POC-CF and 5.7% of POC-NK samples. Compiled results from this study and relevant case series through a literature review demonstrated an abnormality detection rate(ADR) of 35% for chromosomal abnormalities in POC-CF samples, 3.7% for pathogenic CNVs in POC-CF samples, and 4.6% for pathogenic CNVs in POC-NK samples. Ingenuity Pathway Analysis(IPA) was performed on the genes from pathogenic CNVs found in POC samples. The denoted primary gene networks suggested that apoptosis and cell proliferation pathways are involved in miscarriage. In summary, a similar spectrum of cytogenomic abnormalities was observed in POC culture success and POC-CF samples. A threshold effect correlating the number of dosage-sensitive genes in a chromosome with the observed frequency of autosomal trisomy is proposed. A rationalized approach using firstly fluorescence in situ hybridization(FISH) testing with probes of chromosomes X/Y/18, 13/21, and 15/16/22 for common aneuploidies and polyploidies and secondly a CGH for other cytogenomic abnormalities is recommended for POC-CF samples.     

Robust Clustering Using Exponential Power Mixtures

Summary Clustering is a widely used method in extracting useful information from gene expression data, where unknown correlation structures in genes are believed to persist even after normalization. Such correlation structures pose a great challenge on the conventional clustering methods, such as the Gaussian mixture (GM) model, k‐means (KM), and partitioning around medoids (PAM), which are not robust against general dependence within data. Here we use the exponential power mixture model to increase the robustness of clustering against general dependence and nonnormality of the data. An expectation–conditional maximization algorithm is developed to calculate the maximum likelihood estimators (MLEs) of the unknown parameters in these mixtures. The Bayesian information criterion is then employed to determine the numbers of components of the mixture. The MLEs are shown to be consistent under sparse dependence. Our numerical results indicate that the proposed procedure outperforms GM, KM, and PAM when there are strong correlations or non‐Gaussian components in the data.     

Normal huntingtin function: an alternative approach to Huntington's disease

Several neurological diseases are characterized by the altered activity of one or a few ubiquitously expressed cell proteins, but it is not known how these normal proteins turn into harmful executors of selective neuronal cell death. We selected huntingtin in Huntington's disease to explore this question because the dominant inheritance pattern of the disease seems to exclude the possibility that the wild-type protein has a role in the natural history of this condition. However, even in this extreme case, there is considerable evidence that normal huntingtin is important for neuronal function and that the activity of some of its downstream effectors, such as brain-derived neurotrophic factor, is reduced in Huntington's disease.     

研究中国人术中喉返神经监测肌电信号的正常值范围

目的:应用术中神经监测技术(intraoperative neuromonitoring IONM),探讨国人喉返神经肌电信号的正常值范围。方法:烟台毓璜顶医院甲状腺外科自2009年11月到2012年1月对300例甲状腺开放手术行术中喉返神经实时监测,术中一根回路电极斜行刺入同侧切口外缘直径约2 cm范围皮肤,两根记录电极斜行刺入环甲肌,手持刺激探针垂直刺激神经,发出"嘟嘟嘟"提示音,同时监测仪显示器显示肌电波形,并记录潜伏期及波幅值。结果:285例(386条)清晰显示肌电波形;15例未引出肌电波形,其中10例因机器故障和麻醉因素造成假阴性,5例因肿瘤浸润环甲肌,无法插入记录电极造成。无永久性喉返神经损伤,暂时性神经损伤2例,术后1月内恢复。V1 95%可信区间482.66～574.28μV,R1 95%可信区间521.85～615.05μV,V2 95%可信区间440.92～531.84μV,R2 95%可信区间489.25～582.05μV,T可信区间19.88～20.12ms。结论:明确喉返神经肌电信号正常值范围,为识别、确认喉返神经及通过肌电信号判断神经功能完整性提供依据,降低喉返神经损伤率,利于判断喉返神经非离端性损伤,并寻找损伤原因。     

Harmonic Mean Approach to Regression Analysis Under Heteroscedasticity and Nonnormality

In this paper, by combining the harmonic mean approach with the Welch and the James procedure (see WELCH 1951, JAMES, 1951), we develop some robust procedures for testing parallelism in several straight lines under heteroscedasticity and nonnormality. Through Monte Carlo simulations it is shown that these new tests are quite robust with respect to departure from normality. For small sample sizes, however, the TAN-TABATABAI (1984) F_β and F*_β tests appear to be more powerful than the new tests, although when sample sizes are not small, there are hardly any differences between the Tan-Tabatabai F_β and F*_β tests and the new tests.     

Changes in the Levels and Forms of Cholinesterases in the Blood Plasma of Normal and Dystrophic Chickens

Abstract: Acetylcholinesterase (AChE) and pseudocholinesterase (°ChE) were analysed in the blood plasma of developing chickens, both normal and those with inherited muscular dystrophy. The amounts and the molecular forms of each were examined. °ChE concentration rises in the plasma of normal and dystrophic chicks at the end of embryonic development and is maintained after hatching at a constant, relatively high level, accounting for 90-95% of total cholinesterase activity in normal plasma. This level is maintained in normal and dystrophic chickens. In embryonic plasma of both normal and dystrophic chicks, on the other hand, the levels of AChE are higher than those of °ChE. Immediately after hatching the AChE level decreases rapidly in normal plasma, reaching a very low level by 2-3 weeks ex ovo. The AChE level in plasma from dystrophic birds, although less than normal from day 19 in ovo to 2 weeks ex ovo, subsequently increases to peak around 4 months at levels 15-20-fold of those in normal birds. There is virtually no enzyme of either type in the erythrocytes of normal or dystrophic chickens. The changes of AChE in plasma were correlated with the alterations of AChE in dystrophic fast-twitch muscles, suggesting that the latter pool is a precursor of the plasma AChE. Both the AChE and the °ChE in plasma exist in multiple molecular forms, which are similar to certain of those found previously in the muscles of these birds. The major form (60-80%) of both enzymes in the plasma is the M form (sedimentation coefficient ≥11 S) in all cases, but it is accompanied by certain other forms. In no case is there any of the heaviest form (H₂, 19-20 S) of AChE or of °ChE found in normal and dystrophic muscle, which is attached at the synapses in normal muscle. The pattern of forms of plasma °ChE is constant at all ages, and in normal and dystrophic chickens. The pattern of forms of AChE in the plasma, in contrast, varies with age and with dystrophy in a characteristic manner. The sedimentation coefficients and the amounts of the enzymes in fast-twitch muscle of dystrophic animals are compared with those of the plasma forms, and an interpretation is given of the characteristic patterns of AChE and of χE in their blood.