Living with observational data in biological anthropology

The establishment of cause and effect relationships is a fundamental objective of scientific research. Many lines of evidence can be used to make cause–effect inferences. When statistical data are involved, alternative explanations for the statistical relationship need to be ruled out. These include chance (apparent patterns due to random factors), confounding effects (a relationship between two variables because they are each associated with an unmeasured third variable), and sampling bias (effects due to preexisting properties of compared groups). The gold standard for managing these issues is a controlled randomized experiment. In disciplines such as biological anthropology, where controlled experiments are not possible for many research questions, causal inferences are made from observational data. Methods that statisticians recommend for this difficult objective have not been widely adopted in the biological anthropology literature. Issues involved in using statistics to make valid causal inferences from observational data are discussed.     

Two‐sample nonparametric likelihood inference based on incomplete data with an application to a pneumonia study

The clinical pulmonary infection score (CPIS) and bronchoalveolar lavage (BAL) are important diagnostic variables of pneumonia for forcefully ventilated patients who are susceptible to nosocomial infection. Because of its invasive nature, BAL is performed for patients only if the CPIS is greater than a certain threshold value. Thus, CPIS and BAL are closely related, yet BAL values are substantially missing. In a randomized clinical trial, the control and oral treatment groups were compared based on the outcomes from these procedures. Because of the relevance of both outcomes with respect to evaluating the efficacy of treatments, we propose and examine a nonparametric test based on these outcomes, which employs the empirical likelihood methodology. While efficient parametric methods are available when data are observed incompletely, performing appropriate goodness‐of‐fit tests to justify the parametric assumptions is difficult. Our motivation is to provide an approach based on no particular distributional assumption, which enables us to use all observed bivariate data, whether completed or not in an approximate likelihood manner. A broad Monte Carlo study evaluates the asymptotic properties and efficiency of the proposed method based on various sample sizes and underlying distributions. The proposed technique is applied to a data set from a pneumonia study demonstrating its practical worth.     

Compliance subsampling designs for comparative research: estimation and optimal planning

For studies with treatment noncompliance, analyses have been developed recently to better estimate treatment efficacy. However, the advantage and cost of measuring compliance data have implications on the study design that have not been as systematically explored. In order to estimate better treatment efficacy with lower cost, we propose a new class of compliance subsampling (CSS) designs where, after subjects are assigned treatment, compliance behavior is measured for only subgroups of subjects. The sizes of the subsamples are allowed to relate to the treatment assignment, the assignment probability, the total sample size, the anticipated distributions of outcome and compliance, and the cost parameters of the study. The CSS design methods relate to prior work (i) on two-phase designs in which a covariate is subsampled and (ii) on causal inference because the subsampled postrandomization compliance behavior is not the true covariate of interest. For each CSS design, we develop efficient estimation of treatment efficacy under binary outcome and all-or-none observed compliance. Then we derive a minimal cost CSS design that achieves a required precision for estimating treatment efficacy. We compare the properties of the CSS design to those of conventional protocols in a study of patient choices for medical care at the end of life.     

Genomic control for association studies

A dense set of single nucleotide polymorphisms (SNP) covering the genome and an efficient method to assess SNP genotypes are expected to be available in the near future. An outstanding question is how to use these technologies efficiently to identify genes affecting liability to complex disorders. To achieve this goal, we propose a statistical method that has several optimal properties: It can be used with case control data and yet, like family-based designs, controls for population heterogeneity; it is insensitive to the usual violations of model assumptions, such as cases failing to be strictly independent; and, by using Bayesian outlier methods, it circumvents the need for Bonferroni correction for multiple tests, leading to better performance in many settings while still constraining risk for false positives. The performance of our genomic control method is quite good for plausible effects of liability genes, which bodes well for future genetic analyses of complex disorders.     

Reduced sensitivity to hidden bias at upper quantiles in observational studies with dilated treatment effects

When a treatment has a dilated effect, with larger effects when responses are higher, there can be much less sensitivity to bias at upper quantiles than at lower quantiles; i.e., small, plausible hidden biases might explain the ostensible effect of the treatment for many subjects, and yet only quite large hidden biases could explain the effect on a few subjects having dramatically elevated responses. An example concerning kidney function of cadmium workers is discussed in detail. In that example, the treatment effect is far from additive: It is plausibly zero at the lower quartile of responses to control, and it is large and fairly insensitive to bias at the upper quartile.     

Small-sample confidence regions in exponential families

This article presents an algorithm for small-sample conditional confidence regions for two or more parameters for any discrete regression model in the generalized linear interactive model family. Regions are constructed by careful inversion of conditional hypothesis tests. This method presupposes the use of approximate or exact techniques for enumerating the sample space for some components of the vector of sufficient statistics conditional on other components. Such enumeration may be performed exactly or by exact or approximate Monte Carlo, including the algorithms of Kolassa and Tanner (1994, Journal of the American Statistical Association 89, 697-702; 1999, Biometrics 55, 246-251). This method also assumes that one can compute certain conditional probabilities for a fixed value of the parameter vector. Because of a property of exponential families, one can use this set of conditional probabilities to directly compute the conditional probabilities associated with any other value of the vector of the parameters of interest. This observation dramatically reduces the computational effort required to invert the hypothesis test to obtain the confidence region. To construct a region with confidence level 1 - alpha, the algorithm begins with a grid of values for the parameters of interest. For each parameter vector on the grid (corresponding to the current null hypothesis), one transforms the initial set of conditional probabilities using exponential tilting and then calculates the p value for this current null hypothesis. The confidence region is the set of parameter values for which the p value is at least alpha.     

Statistical inference for serial dilution assay data

Serial dilution assays are widely employed for estimating substance concentrations and minimum inhibitory concentrations. The Poisson-Bernoulli model for such assays is appropriate for count data but not for continuous measurements that are encountered in applications involving substance concentrations. This paper presents practical inference methods based on a log-normal model and illustrates these methods using a case application involving bacterial toxins.     

Using molecular markers with high mutation rates to obtain estimates of relative population size and to distinguish the effects of gene flow and mutation: a demonstration using data from endemic Mauritian skinks

We propose a method of analysing genetic data to obtain separate estimates of the size (N(p)) and migration rate (m(p)) for the sampled populations, without precise prior knowledge of mutation rates at each locus ( micro(L)). The effects of migration and mutation can be distinguished because high migration has the effect of reducing genetic differentiation across all loci, whereas a high mutation rate will only affect the locus in question. The method also takes account of any differences between the spectra of immigrant alleles and of new mutant alleles. If the genetic data come from a range of population sizes, and the loci have a range of mutation rates, it is possible to estimate the relative sizes of the different N(p) values, and likewise the m(p) and the micro(L). Microsatellite loci may also be particularly appropriate because loci with a high mutation rate can reach mutation-drift-migration equilibrium more quickly, and because the spectra of mutants arriving in a population can be particularly distinct from the immigrants. We demonstrate this principle using a microsatellite data set from Mauritian skinks. The method identifies low gene flow between a putative new species and populations of its sister species, whereas the differentiation of two other populations is attributed to small population size. These distinct interpretations were not readily apparent from conventional measures of genetic differentiation and gene diversity. When the method is evaluated using simulated data sets, it correctly distinguishes low gene flow from small population size. Loci that are not at mutation-migration-drift equilibrium can distort the parameter estimates slightly. We discuss strategies for detecting and overcoming this effect.     

Statistical analysis of noninferiority trials with a rate ratio in small-sample matched-pair designs

Testing of noninferiority has become increasingly important in modern medicine as a means of comparing a new test procedure to a currently available test procedure. Asymptotic methods have recently been developed for analyzing noninferiority trials using rate ratios under the matched-pair design. In small samples, however, the performance of these asymptotic methods may not be reliable, and they are not recommended. In this article, we investigate alternative methods that are desirable for assessing noninferiority trials, using the rate ratio measure under small-sample matched-pair designs. In particular, we propose an exact and an approximate exact unconditional test, along with the corresponding confidence intervals based on the score statistic. The exact unconditional method guarantees the type I error rate will not exceed the nominal level. It is recommended for when strict control of type I error (protection against any inflated risk of accepting inferior treatments) is required. However, the exact method tends to be overly conservative (thus, less powerful) and computationally demanding. Via empirical studies, we demonstrate that the approximate exact score method, which is computationally simple to implement, controls the type I error rate reasonably well and has high power for hypothesis testing. On balance, the approximate exact method offers a very good alternative for analyzing correlated binary data from matched-pair designs with small sample sizes. We illustrate these methods using two real examples taken from a crossover study of soft lenses and a Pneumocystis carinii pneumonia study. We contrast the methods with a hypothetical example.