Comparison of statistical significance criteria

We study and compare two classes of statistical criteria to assess the significance of exceptional words. Indeed, the Z-score-like criteria, or the normal approximation that is a strict equivalent, suffer from several drawbacks in terms of sensitivity and specificity. Thanks to the combinatorial structure of words, a computation of the exact P-value has been made possible by recent mathematical results. We study here the drawbacks of the Z-score, the choice of the threshold and the tightness to the P-value. A major conclusion is that the normal approximation is always very poor and overestimates statistical significance.     

The cluster method in conducting epidemiological research

The Expanded Programme on Immunization (EPI) whose goal is to reduce morbidity and mortality by providing children with immunizations against diphtheria, pertussis, tetanus, poliomyelitis, measles, and tuberculosis continually faces the problem of documenting immunization coverage rates. Therefore the EPI seeks simple, effective, and inexpensive methods of evaluation which could be implemented in different countries. An example of such a method is a simplified cluster sampling technique of estimation of immunization coverage through the examination of 210 children, selected randomly as 30 groups of 7 children each. In 1978-1984 more than 1000 immunization coverage surveys were performed all over the world, mainly in developing countries. In a modified way this method is also used to collect data on morbidity and mortality of certain EPI target diseases as well as diarrhoeal diseases.     

Radiation hybrid mapping: statistical criteria and their properties

Statistical tests specially developed for radiation hybrid (RH) mapping, modified likelihood-ratio test and Akaike's information content, are proposed. They allow several most likely gene orders to be identified to a certain accuracy. The properties of the proposed tests that characterize the accuracy of gene ordering as dependent on the number of genes mapped, the position of the selective gene, and the sample size have been studied. Recommendations as to the practical use of these tests are presented.     

Ethnicity as a variable in epidemiological research.

Ethnicity is used increasingly as a key variable to describe health data, and ethnic monitoring in the NHS will further stimulate this trend. We identify four fundamental problems with ethnicity in this type of research: the difficulties of measurement, the heterogeneity of the populations being studied, lack of clarity about the research purpose of the research, and ethnocentricity affecting the interpretation and use of data. Ethnicity needs to be used carefully to be a useful tool for health research. We make nine recommendations for future practice, one of which is that ethnicity and race should be recognised and treated as distinct concepts.     

Using statistical equivalence testing in clinical biofeedback research

The technique of statistical equivalence testing is described and recommended for use in clinical biofeedback research. Equivalence testing is valuable in aiding the interpretation of negative results or statistically significant results where effects are small in clinical terms. The method is also useful for establishing the similarity of treatment groups at baseline or for showing that the effect of a potentially confounding variable is tolerably small. Finally, equivalence testing is recommended as a method for documenting the equivalence of biofeedback therapy to proven conventional medical therapies in clinical equivalence trials. Examples, drawn from published literature, are provided.     

An application of statistical models in mental health research

Two statistical models are presented to describe the admission-discharge process of a psychiatric unit. Both models have the form of Markov processes. Common statistical terms such as length of stay in hospital, time off books, patients on books, and number of admissions are related to characteristics of the models. The models permit an assessment of the effect of studying statistical data based on cohorts of individuals rather than individuals considered separately. In addition, with refinement, it is expected that these models will permit the use of more sophisticated statistical methods in psychiatric research problems. The authors wish to acknowledge the continuing support of the Psychiatric Services Branch, Saskatchewan Department of Health and National Health Grant 607-7-129.     

The formulation of epidemiological hypotheses by using the methods of multivariate statistical analysis

The work demonstrates the main approaches to the use of the methods of multidimensional analysis for the creation of a hypothesis on the mechanism of the epidemiological process of dysentery in organized groups. The main risk factors have been established, and their role in the formation of annual, all-the-year-round and seasonal dysentery morbidity has been quantitatively evaluated. The results of analysis show the existence of diverse variants of the alimentary route of the transmission of infection, maintaining the epidemic process of dysentery, and the necessity of differentiating measures for the prophylaxis of all-the-year-round and seasonal morbidity.     

Model of Grapholita molesta spring emergence in pear orchards based on statistical information criteria

The oriental fruit moth, Grapholita molesta, is becoming a large threat to Korean pear production. Timely management of the egg and early larval stages from the spring emergence is critical to reduce the G. molesta population during the pear growing season. A model was developed to precisely predict the spring occurrence of G. molesta adults as a function of accumulated degree-days. The model was validated with male moth caught in sex pheromone-baited traps placed in pear orchards at two major pear production regions (Icheon and Naju) of Korea in 2010. We applied nine distribution models to describe the cumulative proportions of G. molesta males caught relative to accumulated degree-days. The observed phenology of the G. molesta spring population was well described by the nine models. The predicted dates for the cumulative 50% male moth catches were within a 5 day period. Based on statistical information criteria (Akaike's and Bayes–Schwartz information criteria), we recommend the sigmoid function referred by Brown and Mayer, because of its ease of use and meaningfulness; the parameter “b” denotes the degree-day accumulation at 50% moth emergence. The G. molesta spring emergence model could be applied to determine optimal chemical treatment timing for controlling G. molesta in fruit tree orchards and further help to develop a full-cycle phenology model of G. molesta.     

A statistical method for idiographic analyses in biogeographical research

Idiographic analyses (i.e. detailed analyses of single species ranges) can be criticized for subjective and speculative reasoning. Medoid partition is suggested here as a method to perform a statistically supported idiographic analysis. The medoid algorithms attempt to group objects into clusters by finding a set of representative objects called medoids. If areas are the objects that are clustered using species occurrences (0/1) as variables, each cluster will be characterized by a medoid area. The species that characterize each medoid are representative of the entire cluster to which the medoid belongs and can be regarded as (statistically supported) species 'characteristic' of the main distributional patterns observed in the study system and can be used to draw idiographic observations. To illustrate the issues involved, the Coleoptera Tenebrionidae of the Aegean Islands (Greece) were analysed. Two species appeared to be characteristic of the Balkan cluster, while eight species were characteristic of the Anatolian one, and two species were equally distributed in both areas. Idiographic considerations based on these species outlined the importance of a Balkano–Anatolian discontinuity in the Aegean that prevented species dispersal between the two landmasses. This study illustrates that medoid analysis may help the researcher to find some representative patterns from a puzzling distribution. Traditional idiographic analyses can be biased by the fact that species are selected ad hoc . Thus, one cannot establish if results are truly objective or if the author intentionally selected, from a wider array of species, those that supported some preferred patterns. Medoid clustering uses the full array of species to find clusters of areas. After clusters are objectively defined, their medoids are examined to find species that mostly contributed to cluster definition, and the distribution patterns of these species are interpreted.     

Two statistical criteria to choose the method for dilution correction in metabolomic urine measurements

Introduction

Different normalization methods are available for urinary data. However, it is unclear which method performs best in minimizing error variance on a certain data-set as no generally applicable empirical criteria have been established so far.

Objectives

The main aim of this study was to develop an applicable and formally correct algorithm to decide on the normalization method without using phenotypic information.

Methods

We proved mathematically for two classical measurement error models that the optimal normalization method generates the highest correlation between the normalized urinary metabolite concentrations and its blood concentrations or, respectively, its raw urinary concentrations. We then applied the two criteria to the urinary ¹H-NMR measured metabolomic data from the Study of Health in Pomerania (SHIP-0; n?=?4068) under different normalization approaches and compared the results with in silico experiments to explore the effects of inflated error variance in the dilution estimation.

Results

In SHIP-0, we demonstrated consistently that probabilistic quotient normalization based on aligned spectra outperforms all other tested normalization methods. Creatinine normalization performed worst, while for unaligned data integral normalization seemed to most reasonable. The simulated and the actual data were in line with the theoretical modeling, underlining the general validity of the proposed criteria.

Conclusions

The problem of choosing the best normalization procedure for a certain data-set can be solved empirically. Thus, we recommend applying different normalization procedures to the data and comparing their performances via the statistical methodology explicated in this work. On the basis of classical measurement error models, the proposed algorithm will find the optimal normalization method.

     

Using statistical analysis for setting process validation acceptance criteria for biotech products

This paper discusses the challenges of setting process validation acceptance criteria for biotech products for cases where using statistical tools is appropriate. Data are analyzed under three different scenarios that are frequently encountered in biotech applications. Scenario A represents the case when a small data set around center point conditions is available for setting acceptance criteria. Scenario B represents the case when a larger data set within normal operation conditions is available for setting acceptance criteria. Scenario C represents the case when a large characterization data set is available for setting acceptance criteria and it is possible to accurately model the impact of operation conditions on performance of the step. Statistical approaches including mean +/- 3SD, tolerance interval analysis, prediction profiler, and Monte Carlo simulation are applied to the different scenarios. Strengths and shortcomings of the different statistical tools are discussed, and the best approach for each scenario is recommended. It is shown that selection of the right statistical approach is a critical first step toward setting appropriate acceptance criteria.