On Generalized Binomial and Negative Binomial Distributions for Dependent Bernoulli Variables

We study the distributions of the random variables S_n and V_r related to a sequence of dependent Bernoulli variables, where S_n denotes the number of successes in n trials and V_r the number of trials necessary to obtain r successes. The purpose of this article is twofold: (1) Generalizing some results on the “nature” of the binomial and negative binomial distributions we show that S_n and V_r can follow any prescribed discrete distribution. The corresponding joint distributions of the Bernoulli variables are characterized as the solutions of systems of linear equations. (2) We consider a specific type of dependence of the Bernoulli variables, where the probability of a success depends only on the number of previous successes. We develop some theory based on new closed-form representations for the probability mass functions of S_n and V_r which enable direct computations of the probabilities.     

Two-scale spatial models for binary data

A spatial lattice model for binary data is constructed from two spatial scales linked through conditional probabilities. A coarse grid of lattice locations is specified, and all remaining locations (which we call the background) capture fine-scale spatial dependence. Binary data on the coarse grid are modelled with an autologistic distribution, conditional on the binary process on the background. The background behaviour is captured through a hidden Gaussian process after a logit transformation on its Bernoulli success probabilities. The likelihood is then the product of the (conditional) autologistic probability distribution and the hidden Gaussian–Bernoulli process. The parameters of the new model come from both spatial scales. A series of simulations illustrates the spatial-dependence properties of the model and likelihood-based methods are used to estimate its parameters. Presence–absence data of corn borers in the roots of corn plants are used to illustrate how the model is fitted.     

Binomial and Negative Binomial Analogues under Correlated Bernoulli Trials

Several statistical applications demand the adoption of models in which the response is binary but the outcomes of different trials exhibit some degree of correlation. Although the independent case is well known and treated even in elementary textbooks, results on correlated Bernoulli trials are hardly found in the literature. Analogues of the binomial and negative binomial distributions are presented in this article when the correlation is of the Markovian type. Probability-generating function, probability mass function, mean, and variance are derived. The analysis allows illustration of a variety of techniques useful in the study of discrete distributions appropriate for second-level probability courses. An example on customer brand switching discussed by Olkin, Glesser, and Derman is presented as illustration.     

Mean success run length

In this paper we consider mean of success run lengths appearing in a sequence of binary trials. We derive the exact and limiting distributions of mean success run length for i.i.d. Bernoulli trials. The exact distribution of the corresponding random variable is also derived for a sequence of Markov-dependent Bernoulli trials. In addition, a combinatorial formula for the distribution of any success run statistic defined on Markov-dependent trials is presented.     

Shortest and longest length of success runs in binary sequences

The shortest and the longest length of success runs statistics in binary sequences are considered. The sequences are arranged on a line or on a circle. Exact probabilities of these statistics are derived, both in closed formulae via combinatorial analysis, as well as recursively. Furthermore, their joint probability distribution function and cumulative distribution function are obtained. The results are developed first for Bernoulli trials (i.i.d. binary sequences), and then they are generalized to the Polya–Eggenberger sampling scheme. For the latter case, the length of the longest success run is related to other success runs statistics and to reliability of consecutive systems.     

Efficacy and toxicity monitoring via Bayesian predictive probabilities in phase II clinical trials

Bayesian monitoring strategies based on predictive probabilities are widely used in phase II clinical trials that involve a single efficacy binary variable. The essential idea is to control the predictive probability that the trial will show a conclusive result at the scheduled end of the study, given the information at the interim stage and the prior beliefs. In this paper, we present an extension of this approach to incorporate toxicity considerations in single-arm phase II trials. We consider two binary endpoints representing response and toxicity of the experimental treatment and define the result as successful at the conclusion of the study if the posterior probability of an high efficacy and that of a small toxicity are both sufficiently large. At any interim look, the Multinomial-Dirichlet distribution provides the predictive probability of each possible combination of future efficacy and toxicity outcomes. It is exploited to obtain the predictive probability that the trial will yield a positive outcome, if it continues to the planned end. Different possible interim situations are considered to investigate the behaviour of the proposed predictive rules and the differences with the monitoring strategies based on posterior probabilities are highlighted. Simulation studies are also performed to evaluate the frequentist operating characteristics of the proposed design and to calibrate the design parameters.

     

Optimal designs for contingent response models with application to toxicity–efficacy studies

We describe a general family of contingent response models. These models have ternary outcomes constructed from two Bernoulli outcomes, where one outcome is only observed if the other outcome is positive. This family is represented in a canonical form which yields general results for its Fisher information. A bivariate extreme value distribution illustrates the model and optimal design results. To provide a motivating context, we call the two binary events that compose the contingent responses toxicity and efficacy. Efficacy or lack thereof is assumed only to be observable in the absence of toxicity, resulting in the ternary response (toxicity, efficacy without toxicity, neither efficacy nor toxicity). The rate of toxicity, and the rate of efficacy conditional on no toxicity, are assumed to increase with dose. While optimal designs for contingent response models are numerically found, limiting optimal designs can be expressed in closed forms. In particular, in the family of four parameter bivariate location-scale models we study, as the marginal probability functions of toxicity and no efficacy diverge, limiting D optimal designs are shown to consist of a mixture of the D optimal designs for each failure (toxicity and no efficacy) univariately. Limiting designs are also obtained for the case of equal scale parameters.     

The Muscatine children's obesity data reanalysed using pattern mixture models

A set of longitudinal binary, partially incomplete, data on obesity among children in the USA is reanalysed. The multivariate Bernoulli distribution is parameterized by the univariate marginal probabilities and dependence ratios of all orders, which together support maximum likelihood inference. The temporal association of obesity is strong and complex but stationary. We fit a saturated model for the distribution of response patterns and find that non-response is missing completely at random for boys but that the probability of obesity is consistently higher among girls who provided incomplete records than among girls who provided complete records. We discuss the statistical and substantive features of, respectively, pattern mixture and selection models for this data set.     

Better approximate confidence intervals for a binomial parameter

This paper discusses five methods for constructing approximate confidence intervals for the binomial parameter Θ, based on Y successes in n Bernoulli trials. In a recent paper, Chen (1990) discusses various approximate methods and suggests a new method based on a Bayes argument, which we call method I here. Methods II and III are based on the normal approximation without and with continuity correction. Method IV uses the Poisson approximation of the binomial distribution and then exploits the fact that the exact confidence limits for the parameter of the Poisson distribution can be found through the x² distribution. The confidence limits of method IV are then provided by the Wilson-Hilferty approximation of the x². Similarly, the exact confidence limits for the binomial parameter can be expressed through the F distribution. Method V approximates these limits through a suitable version of the Wilson-Hilferty approximation. We undertake a comparison of the five methods in respect to coverage probability and expected length. The results indicate that method V has an advantage over Chen's Bayes method as well as over the other three methods.     

Adaptive Percolation Using Subjective Likelihoods

A phenomenon that I call “adaptive percolation” commonly arises in biology, business, economics, defense, finance, manufacturing, and the social sciences. Here one wishes to select a handful of entities from a large pool of entities via a process of screening through a hierarchy of sieves. The process is not unlike the percolation of a liquid through a porous medium. The probability model developed here is based on a nested and adaptive Bayesian approach that results in the product of beta-binomial distributions with common parameters. The common parameters happen to be the observed data. I call this the percolated beta-binomial distribution . The model turns out to be a slight generalization of the probabilistic model used in percolation theory. The generalization is a consequence of using a subjectively specified likelihood function to construct a probability model. The notion of using likelihoods for constructing probability models is not a part of the conventional toolkit of applied probabilists. To the best of my knowledge, a use of the product of beta-binomial distributions as a probability model for Bernoulli trials appears to be new. The development of the material of this article is illustrated via data from the 2009 astronaut selection program, which motivated this work.     

Forecasting future values of changing sequences

Sequences of independent random variables are observed and on the basis of these observations future values of the process are forecast. The Bayesian predictive density of k future observations for normal, exponential, and binomial sequences which change exactly once are analyzed for several cases. It is seen that the Bayesian predictive densities are mixtures of standard probability distributions. For example, with normal sequences the Bayesian predictive density is a mixture of either normal or t-distributions, depending on whether or not the common variance is known. The mixing probabilities are the same as those occurring in the corresponding posterior distribution of the mean(s) of the sequence. The predictive mass function of the number of future successes that will occur in a changing Bernoulli sequence is computed and point and interval predictors are illustrated.     

Binary Time Series Modeling with Application to Adhesion Frequency Experiments

Repeated adhesion frequency assay is the only published method for measuring the kinetic rates of cell adhesion. Cell adhesion plays an important role in many physiological and pathological processes. Traditional analysis of adhesion frequency experiments assumes that the adhesion test cycles are independent Bernoulli trials. This assumption can often be violated in practice. Motivated by the analysis of repeated adhesion tests, a binary time series model incorporating random effects is developed in this paper. A goodness-of-fit statistic is introduced to assess the adequacy of distribution assumptions on the dependent binary data with random effects. The asymptotic distribution of the goodness-of-fit statistic is derived and its finite-sample performance is examined via a simulation study. Application of the proposed methodology to real data from a T-cell experiment reveals some interesting information, including the dependency between repeated adhesion tests.     

On binomial and circular binomial distributions of orderk forl-overlapping success runs of lengthk

The number ofl-overlapping success runs of lengthk inn trials, which was introduced and studied recently, is presently reconsidered in the Bernoulli case and two exact formulas are derived for its probability distribution function in terms of multinomial and binomial coefficients respectively. A recurrence relation concerning this distribution, as well as its mean, is also obtained. Furthermore, the number ofl-overlapping success runs of lengthk inn Bernoulli trials arranged on a circle is presently considered for the first time and its probability distribution function and mean are derived. Finally, the latter distribution is related to the first, two open problems regarding limiting distributions are stated, and numerical illustrations are given in two tables. All results are new and they unify and extend several results of various authors on binomial and circular binomial distributions of orderk.     

A Bayesian analysis of ratios of proportions in tree-structured experiments

This paper concerns the calculation of Bayes estimators of ratios of outcome proportions generated by the replication of an arbitrary tree-structured compound Bernoulli experiment under a multinomial-type sampling scheme. Here the compound Bernoulli experiment is treated as a collection of linear sequences of independent generalized Bernoulli trials having Dirichlet type 1 prior probability distributions. A method of obtaining a closed-form expression of the cumulative distribution function of the ratio of proportions – from its Meijer G-function representation – is described. Bayes point and interval estimators are directly obtained from the properties the distribution function as well as its related probability density function. In addition, the density function is used to derive the probability mass function of the predictive distribution any two associated outcome categories of the experiment – under an inverse multinomial-type sampling scheme. An illustrative numerical example concerning a Bayesian analysis of a simple tree-structured mortality model for medical patients who have suffered an acute myocardial infarction (heart attack) is also included.     

Bootstrap goodness-of-fit test for the beta-binomial model

A common question in the analysis of binary data is how to deal with overdispersion. One widely advocated sampling distribution for overdispersed binary data is the beta-binomial model. For example, this distribution is often used to model litter effects in toxicological experiments. Testing the null hypothesis of a beta-binomial distribution against all other distributions is difficult, however, when the litter sizes vary greatly. Herein, we propose a test statistic based on combining Pearson statistics from individual litter sizes, and estimate the p-value using bootstrap techniques. A Monte Carlo study confirms the accuracy and power of the test against a beta-binomial distribution contaminated with a few outliers. The method is applied to data from environmental toxicity studies.     

A bivariate gamma mixture distribution

In this paper we construct a bivariate gamma mixture distribution by allowing the scale parameters of the two marginals to have a generalized Bernoulli distribution. We study the statistical properties of this distribution and discuss the estimation of the parameters. The distribution is then fitted to two bivariate data sets. Data on the age and lactation period of cows are described well by the proposed model, but it fails to fit Johansen's bean data properly because of theoretical constraints on the correlation.     

Exact tests of binomial p=1/2 when one binomial event possibly has different probability

The data are n independent random binomial events, each resulting in success or failure. The event outcomes are believed to be trials from a binomial distribution with success probability p, and tests of p=1/2 are desired. However, there is the possibility that some unidentified event has a success probability different from the common value p for the other n?1 events. Then, tests of whether this common p equals 1/2 are desired. Fortunately, two-sided tests can be obtained that simultaneously are applicable for both situations. That is, the significance level for a test is same when one event has a different probability as when all events have the same probability. These tests are the usual equal-tail tests for p=1/2 (based on n independent trials from a binomial distribution).     

On the heterogeneity of proportions

The comparison amongmproportions can be viewed as the clustering of the means of Bernoulli trials. By introducing a distribution which is supported on the means of Bernoulli trials, we suggest a moment method approach to determine the center of the clusters. We also suggest using model selection criteria rather than the usual testing hypothesis approach to determine the grouping of the means. The discrepancy function for all possible models are compared based on the bootstrap results.     

Simulating Media Exposure Distributions

We examine three media exposure distribution (e.d.) simulation methods. The first is based on the maximum likelihood estimate of an individual's exposure, the second on ‘personal probability’ (Greene 1970) and the third on a dependent Bernoulli trials model (Klotz 1973). The last method uses population exposure probabilities rather than individual exposure probabilities, thereby markedly reducing computation time. Magazine exposure data are used to compare the accuracy and computation times of the simulation methods with a log–linear e.d. model (Danaher 1988b) and the popular Metheringham (1964) model based on the beta–binomial distribution (BBD). The results show that the simulation methods are not as accurate as the log– linear model but are more accurate than Metheringham's model, However, all the simulation methods take less computation time than the log–linear model for schedules with more than six magazines, making them viable competitors for large schedule sizes     

Waiting time distributions for a run with additional constraints

This paper proposes a method for obtaining the exact probability of occurrence of the first success run of specified length with the additional constraint that at every trial until the occurrence of the first success run the number of successes up to the trial exceeds that of failures. For the sake of the additional constraint, the problem cannot be solved by the usual method of conditional probability generating functions. An idea of a kind of truncation is introduced and studied in order to solve the problem. Concrete methods for obtaining the probability in the cases of Bernoulli trials and time-homogeneous {0,1}$\{0,1\}$-valued Markov dependent trials are given. As an application of the results, a modification of the start-up demonstration test is studied. Numerical examples which illustrate the feasibility of the results are also given.