A sequential design for maximizing the probability of a favourable response

Consider the situation in which subjects arrive sequentially for a treatment and in which there are two distinct ways in which the treatment may fail. Treatments are given at different dosages, and the probabilities of the two failure types vary with dose. Assuming that decreasing the chances of one failure type increases the chances of the other, we say the failures oppose each other. Also assume that one failure type is primary in that, if it occurs, it censors the trial, so that observation of the secondary failure type is contingent on no failure of the primary type. We are interested in designs that provide information about the dose that maximizes the probability of success, i.e., the optimal dose, while treating very few subjects at dosages that have high risks of failure. Assuming that dosages belong to a discrete set, we show that a randomized version of the Pólya urn scheme causes dose selection to be progressively biased so as to favour those doses that produce success with higher probability.     

An analysis of pulsation periods of long period variable stars

Summary.  We report the results of a period change analysis of time series observations for 378 pulsating variable stars. The null hypothesis of no trend in expected periods is tested for each of the stars. The tests are non-parametric in that potential trends are estimated by local linear smoothers. Our testing methodology has some novel features. First, the null distribution of a test statistic is defined to be the distribution that results in repeated sampling from a population of stars. This distribution is estimated by means of a bootstrap algorithm that resamples from the collection of 378 stars. Bootstrapping in this way obviates the problem that the conditional sampling distribution of a statistic, given a particular star, may depend on unknown parameters of that star. Another novel feature of our test statistics is that one-sided cross-validation is used to choose the smoothing parameters of the local linear estimators on which they are based. It is shown that doing so results in tests that are tremendously more powerful than analogous tests that are based on the usual version of cross-validation. The positive false discovery rate method of Storey is used to account for the fact that we simultaneously test 378 hypotheses. We ultimately find that 56 of the 378 stars have changes in mean pulsation period that are significant when controlling the positive false discovery rate at the 5% level.     

Comparing mean ranks for repeated measures data

Rank tests are considered that compare t treatments in repeated measures designs. A statistic is given that contains as special cases several that have been proposed for this problem, including one that corresponds to the randomized block ANOVA statistic applied to the rank transformed data. Another statistic is proposed, having a null distribution holding under more general conditions, that is the rank transform of the Hotelling statistic for repeated measures. A statistic of this type is also given for data that are ordered categorical rather than fully rankedo Unlike the Friedman statistic, the statistics discussed in this article utilize a single ranking of the entire sample. Power calculations for an underlying normal distribution indicate that the rank transformed ANOVA test can be substantially more powerful than the Friedman test.     

Heterogeneity and study size in random-effects meta-analysis

It is well known that heterogeneity between studies in a meta-analysis can be either caused by diversity, for example, variations in populations and interventions, or caused by bias, that is, variations in design quality and conduct of the studies. Heterogeneity that is due to bias is difficult to deal with. On the other hand, heterogeneity that is due to diversity is taken into account by a standard random-effects model. However, such a model generally assumes that heterogeneity does not vary according to study-level variables such as the size of the studies in the meta-analysis and the type of study design used. This paper develops models that allow for this type of variation in heterogeneity and discusses the properties of the resulting methods. The models are fitted using the maximum-likelihood method and by modifying the Paule–Mandel method. Furthermore, a real-world argument is given to support the assumption that the inter-study variance is inversely proportional to study size. Under this assumption, the corresponding random-effects method is shown to be connected with standard fixed-effect meta-analysis in a way that may well appeal to many clinicians. The models and methods that are proposed are applied to data from two large systematic reviews.     

Estimating numbers of infectious units from serial dilution assays

Summary.  The paper concerns the design and analysis of serial dilution assays to estimate the infectivity of a sample of tissue when it is assumed that the sample contains a finite number of indivisible infectious units such that a subsample will be infectious if it contains one or more of these units. The aim of the study is to estimate the number of infectious units in the original sample. The standard approach to the analysis of data from such a study is based on the assumption of independence of aliquots both at the same dilution level and at different dilution levels, so that the numbers of infectious units in the aliquots follow independent Poisson distributions. An alternative approach is based on calculation of the expected value of the total number of samples tested that are not infectious. We derive the likelihood for the data on the basis of the discrete number of infectious units, enabling calculation of the maximum likelihood estimate and likelihood-based confidence intervals. We use the exact probabilities that are obtained to compare the maximum likelihood estimate with those given by the other methods in terms of bias and standard error and to compare the coverage of the confidence intervals. We show that the methods have very similar properties and conclude that for practical use the method that is based on the Poisson assumption is to be recommended, since it can be implemented by using standard statistical software. Finally we consider the design of serial dilution assays, concluding that it is important that neither the dilution factor nor the number of samples that remain untested should be too large.     

Adaptive nonparametric density estimation with missing observations

It is well known that if some observations in a sample from the probability density are not available, then in general the density cannot be estimated. A possible remedy is to use an auxiliary variable that explains the missing mechanism. For this setting a data-driven estimator is proposed that mimics performance of an oracle that knows all observations from the sample. It is also proved that the estimator adapts to unknown smoothness of the density and its mean integrated squared error converges with a minimax rate. A numerical study, together with the analysis of a real data, shows that the estimator is feasible for small samples.     

The Interestingness Paradox in Pattern Discovery

Noting that several rule discovery algorithms in data mining can produce a large number of irrelevant or obvious rules from data, there has been substantial research in data mining that addressed the issue of what makes rules truly 'interesting'. This resulted in the development of a number of interestingness measures and algorithms that find all interesting rules from data. However, these approaches have the drawback that many of the discovered rules, while supposed to be interesting by definition, may actually (1) be obvious in that they logically follow from other discovered rules or (2) be expected given some of the other discovered rules and some simple distributional assumptions. In this paper we argue that this is a paradox since rules that are supposed to be interesting, in reality are uninteresting for the above reason. We show that this paradox exists for various popular interestingness measures and present an abstract characterization of an approach to alleviate the paradox. We finally discuss existing work in data mining that addresses this issue and show how these approaches can be viewed with respect to the characterization presented here.     

Improvement over bayes prediction in small samples in the presence of model uncertainty

In an online prediction context, the authors introduce a new class of mongrel criteria that allow for the weighing of candidate models and the combination of their predictions based both on model‐based and empirical measures of their performance. They present simulation results which show that model averaging using the mongrel‐derived weights leads, in small samples, to predictions that are more accurate than that obtained by Bayesian weight updating, provided that none of the candidate models is too distant from the data generator.     

Some practical issues in the evaluation of heterogeneous labour market programmes by matching methods

Summary.  Recently several studies have analysed active labour market policies by using a recently proposed matching estimator for multiple programmes. Since there is only very limited practical experience with this estimator, this paper checks its sensitivity with respect to issues that are of practical importance in this kind of evaluation study. The estimator turns out to be fairly robust to several features that concern its implementation. Furthermore, the paper demonstrates that the matching approach per se is no panacea for solving all the problems of evaluation studies, but that its success depends critically on the information that is available in the data. Finally, a comparison with a bootstrap distribution provides some justification for using a simplified approximation of the distribution of the estimator that ignores its sequential nature.     

Joint modelling of pre-randomisation event counts and multiple post-randomisation survival times with cure rates: application to data for early epilepsy and single seizures

In this paper, we consider the analysis of recurrent event data that examines the differences between two treatments. The outcomes that are considered in the analysis are the pre-randomisation event count and post-randomisation times to first and second events with associated cure fractions. We develop methods that allow pre-randomisation counts and two post-randomisation survival times to be jointly modelled under a Poisson process framework, assuming that outcomes are predicted by (unobserved) event rates. We apply these methods to data that examine the difference between immediate and deferred treatment policies in patients presenting with single seizures or early epilepsy. We find evidence to suggest that post-randomisation seizure rates change at randomisation and following a first seizure after randomisation. We also find that there are cure rates associated with the post-randomisation times to first and second seizures. The increase in power over standard survival techniques, offered by the joint models that we propose, resulted in more precise estimates of the treatment effect and the ability to detect interactions with covariate effects.     

Two concepts of conditionality

Birnbaum's proof that C and M imply L, would lose its force if it is shown that in some situations M is not acceptable. Godambe (1979) has shown that Birnbaum's M is not as obvious or intuitive as the concept that a ‘mere relabelling’ of sample points should make no difference to the interference that can appropriately be drawn from a particular outcome of a given experiment. Akaike (1982) has shown that in certain situations M amounts to the assertion that a relabelling of sample points involving a false reporting of the outcome of an experiment should make no difference to the inference drawn from a particular outcome of a given experiment. It is shown in this paper that in the situation discussed by Akaike, even if M were to be considered acceptable, it is only a modified conditionality principle C? and M which can formally imply L; Birnbaum's conditionality principle C and M do not imply L.     

中国制造还具有劳动力成本优势吗

本文利用2002-2011年的制造业数据核算比较了中国与主要竞争国家的制造业单位劳动力成本,发现我国单位劳动力成本已经高于印尼、泰国和马来西亚等东南亚国家。具体地,东部地区单位劳动力成本2002年以后高于印尼,2007年超过我国中西部地区和泰国,2011年超过马来西亚;中西部地区单位劳动力成本2002年以后高于印尼,2009年超过泰国,2011年超过马来西亚,主要原因在于我国小时劳动力成本上升过快。因而,与印尼、泰国、马来西亚等东南亚国家相比,我国已不具有劳动力成本优势,再加上近年来我国人口红利的消失,中西部地区劳动力成本也已高于东南亚国家,因而,可以解释我国中西部地区为何没能及时接收东部地区的产业转移。     

Information and capacities

Most economists consider that the cases of negative information value that non-Bayesian decision makers seem to exhibit, clearly show that these models are not models representing rational behaviour. We consider this issue for Choquet Expected Utility maximizers in a simple framework, that is the problem of choosing on which event to bet. First, we find a necessary condition to prevent negative information vlaue that we call Separative Monotonicity. This is a weaker condition than Savage Sure thing Principle and it appears that necessity and possibility measures satisfy it and that we cand find conditioning rules such that the information value is always positive. In a second part, we question the way information value is usually measured and suggest that negative information values are merely resulting from an inadequate formula. Yet, we suggest to impose what appears as a weaker requirement, that is, the betting strategy should not be Statistically Dominated. We show for classical updating rules applied to belief functions that this requirement is violated. We consider a class of conditioning rules and exhibit a necessary and sufficient condition in order to satisfy the Statistical Dominance criterion in the case of belief functions. Received: November 2000; revised version: July 2001     

On optimal tests for separate hypotheses and conditional probability integral transformations

Consider the problem of testing the composite null hypothesis that a random sample X₁,…,X_n is from a parent which is a member of a particular continuous parametric family of distributions against an alternative that it is from a separate family of distributions. It is shown here that in many cases a uniformly most powerful similar (UMPS) test exists for this problem, and, moreover, that this test is equivalent to a uniformly most powerful invariant (UMPI) test. It is also seen in the method of proof used that the UMPS test statistic Is a function of the statistics U₁,…,U_n?k obtained by the conditional probability integral transformations (CPIT), and thus that no Information Is lost by these transformations, It is also shown that these optimal tests have power that is a nonotone function of the null hypothesis class of distributions, so that, for example, if one additional parameter for the distribution is assumed known, then the power of the test can not lecrease. It Is shown that the statistics U₁, …, U_n?k are independent of the complete sufficient statistic, and that these statistics have important invariance properties. Two examples at given. The UMPS tests for testing the two-parameter uniform family against the two-parameter exponential family, and for testing one truncation parameter distribution against another one are derived.     

Randomization tests for small samples: an application for genetic expression data

Summary. An advantage of randomization tests for small samples is that an exact P -value can be computed under an additive model. A disadvantage with very small sample sizes is that the resulting discrete distribution for P -values can make it mathematically impossible for a P -value to attain a particular degree of significance. We investigate a distribution of P -values that arises when several thousand randomization tests are conducted simultaneously using small samples, a situation that arises with microarray gene expression data. We show that the distribution yields valuable information regarding groups of genes that are differentially expressed between two groups: a treatment group and a control group. This distribution helps to categorize genes with varying degrees of overlap of genetic expression values between the two groups, and it helps to quantify the degree of overlap by using the P -value from a randomization test. Moreover, a statistical test is available that compares the actual distribution of P -values with an expected distribution if there are no genes that are differentially expressed. We demonstrate the method and illustrate the results by using a microarray data set involving a cell line for rheumatoid arthritis. A small simulation study evaluates the effect that correlated gene expression levels could have on results from the analysis.     

A Note on the Moments and Computer Generation of the Shifted Gompertz Distribution

In this paper we give a class of row-column designs with the property that the i-th row and the j-th column have precisely r treatments in common. A conjecture that such designs are quasi-factorial is disproved by showing that the designs given in this paper are not quasi-factorial. It is also shown that the designs given here are nearly optimal.     

A formulation of the concept of probability based on the use of experimental devices

This paper sets out to identify the abilities that a person needs to be able to successfully use an experimental device, such as a probability wheel or balls in an urn, for the elicitation of subjective probabilities. It is assumed that the successful use of the device requires that the person elicits unique probability values that obey the standard probability laws. This leads to a definition of probability based on the idea of the similarity between the likeliness of events and this concept is naturally extended to the idea that probabilities have strengths, which relates to information about the likeliness of an event that lies beyond a simple probability value. The latter notion is applied to the problem of explaining the Ellsberg paradox. To avoid the definition of probability being circular, probabilities are defined such that they depend on the choice of a reference set of events R which, in simple cases, corresponds to the raw outcomes produced by using an experimental device. However, it is shown that even when the events in R are considered as having an “equal chance” of occurring, the values and/or strengths of probabilities can still be affected by the choice of the set R.     

Standard errors for EM estimation

The EM algorithm is a popular method for computing maximum likelihood estimates. One of its drawbacks is that it does not produce standard errors as a by-product. We consider obtaining standard errors by numerical differentiation. Two approaches are considered. The first differentiates the Fisher score vector to yield the Hessian of the log-likelihood. The second differentiates the EM operator and uses an identity that relates its derivative to the Hessian of the log-likelihood. The well-known SEM algorithm uses the second approach. We consider three additional algorithms: one that uses the first approach and two that use the second. We evaluate the complexity and precision of these three and the SEM in algorithm seven examples. The first is a single-parameter example used to give insight. The others are three examples in each of two areas of EM application: Poisson mixture models and the estimation of covariance from incomplete data. The examples show that there are algorithms that are much simpler and more accurate than the SEM algorithm. Hopefully their simplicity will increase the availability of standard error estimates in EM applications. It is shown that, as previously conjectured, a symmetry diagnostic can accurately estimate errors arising from numerical differentiation. Some issues related to the speed of the EM algorithm and algorithms that differentiate the EM operator are identified.