Prediction intervals in exponential families

A method for obtaining prediction intervals for an outcome of a future experiment is presented. The method uses hypothesis testing as a tool to derive prediction intervals and assumes that the probability distributions of informative and future experiments are one parameter exponential families. Asymptotic similar mean coverage prediction intervals are derived using the score test as a test statistics. Examples are presented and asymptotic prediction limits are compared with the prediction limits given in the literature.     

Towards a job competency profile for agricultural extension instructors: A survey of views of experts

A project is carried out for instructors in the agricultural extension service in Esfahan. Part of the project is study, conducted to inform the process of competency profile development for the instructors. The study, reported in this contribution, focuses on views of experts about competencies needed by the instructors. The survey is based on the methodology for developing models of HRD, in which 257 key experts participated. A mix of methods was used for collecting data. The most important competencies found in this study are subject matter and business understanding, and presentation skill. The most essential outputs are presentation of instructional materials, feedback to learners and equipping learners with new competencies after the course. In this contribution, competencies and outputs are elaborated. Further research is needed regarding performance improvement as a result of the efforts of the instructors, and the views of other stakeholders on the development of the agri-clusters.     

Transfer pricing—A survey using discriminant analysis

There is a considerable literature on transfer pricing. One strand of this literature is empirical and is concerned with the extent to which companies use the various transfer pricing methods. The other strand is strongly normative being directed to devising transfer pricing methods and defining the circumstances under which it is appropriate to use them. This study attempts to address both these questions by using discriminant analysis methods to predict which approach a company will use to determine transfer prices given its weightings of certain ‘determinants’ which theory suggests should be relevant to the decision. Domestic and international transfer pricing are considered separately. Discriminant analysis is shown to be quite successful in predicting the transfer pricing systems adopted by the UK companies surveyed.     

Nonparametric Prediction Intervals for Future Order Statistics in a Proportional Hazard Model

Consider k independent random samples with different sample sizes such that the ith sample comes from the cumulative distribution function (cdf) F _i  = 1 ? (1 ? F)^{α _i} , where α _i is a known positive constant and F is an absolutely continuous cdf. Also, suppose that we have observed the maximum and minimum of the first k samples. This article shows how one can construct the nonparametric prediction intervals for the order statistics of the future samples on the basis of these information. Three schemes are studied and in each case exact expressions for the prediction coefficients of prediction intervals are derived. Numerical computations are given for illustrating the results. Also, a comparison study is done while the complete samples are available.     

Bayesian reconstruction of the missing failure times in exponential distribution

In this paper, we consider the situations in which a middle part of the sample is lost or removed from the experiment. The left-censored sample and Type-II right-censored sample are special cases of this scheme. Point and interval reconstructors for lost failure times are obtained, while the underlying distribution is exponential. To illustrate the performance of the proposed procedures, two numerical examples based on the real data sets are presented. In the case of two-parameter, a simulation study has been done and also the sensitivity of the results to deviations from the true distribution is investigated.     

Nonparametric bootstrapping for hierarchical data

Nonparametric bootstrapping for hierarchical data is relatively underdeveloped and not straightforward: certainly it does not make sense to use simple nonparametric resampling, which treats all observations as independent. We have provided some resampling strategies of hierarchical data, proved that the strategy of nonparametric bootstrapping on the highest level (randomly sampling all other levels without replacement within the highest level selected by randomly sampling the highest levels with replacement) is better than that on lower levels, analyzed real data and performed simulation studies.     

What influences respondents to behave consistently when asked to consent to health record linkage on repeat occasions?

This study constitutes the first longitudinal exploration of consent to link survey and administrative data. It examines variations in consent over time and explores the influence of the respondents’ characteristics (both observed and latent) and the impact of the interviewers on consent co-operation. Respondent inclination to consent is modelled as a latent construct. Most respondents behave consistently over time. However, this consistency is not driven by a strong inclination to consent but rather by the circumstances of the respondents at the time of the interview and by the impact of the interviewers themselves. The findings also show that the change in consent behaviour over time is a clear indication that consent should be treated as a dynamic phenomenon at the individual level.     

Reconstruction of the past lower record values in a proportional reversed hazard rate model

In this paper, the situation in which some lower records from a proportional reversed hazard rate model (PRHRM) are lost at the beginning of the experiment is considered. The reconstruction problem of the past lower records based on observed records from a PRHRM is discussed. Several various methods are used to obtain point reconstructors. More details are given for the Fréchet distribution. Three reconstruction intervals are also obtained and compared in the sense of the expected width for the case of Fréchet distribution. A data set representing the annual flood loss is applied to illustrate the proposed procedure in this paper. Eventually, some concluding remarks are presented.     

Variation within households in consent to link survey data to administrative records: evidence from the UK Millennium Cohort Study

This study expands our knowledge of consent in linking survey and administrative data by studying respondents’ behaviour when consenting to link their own records and when consenting to link those of their children. It develops and tests a number of hypothesised mechanisms of consent, some of which were not explored in the past. The hypotheses cover: parental pride, privacy concerns, loyalty to the survey, pre-existing relations with the agency holding the data, and interviewer effects. The study uses data from the longitudinal Millennium Cohort Study to analyse the correlates of consent in multiple domains (i.e. linkage of education, health and economic records). The findings show that respondent’s behaviour vary depending on the consent domain and on the person within the household for whom consent is sought. In particular, the cohort member’s cognitive skills and the main respondent’s privacy concerns have differential effects on consent. On the other hand, loyalty to the survey proxied by the longitudinal response history has a significant and strong impact on consent irrespective of the outcome. The findings also show that interviewers account for a large proportion of variations in consent even after controlling for the characteristics of the interviewer’s assignment area. In total, it is possible to conclude that the significant impact of some of the correlates will lead to sample bias which needs to be accounted for when working with linked survey and administrative data.     

Influence diagnostics in constrained general linear models

ABSTRACT

Constrained general linear models (CGLMs) have wide applications in practice. Similar to other data analysis, the identification of influential observations that may be potential outliers is an important step beyond in the CGLMs. We develop multiple case-deletion diagnostics for detecting influential observations in the CGLMs. The diagnostics are functions of basic building blocks: studentized residuals, error contrast matrix, and the inverse of the response variable covariance matrix. The basic building blocks are computed only once from the complete data analysis and provide information on the influence of the data on different aspects of the model fit. Computational formulas are given which make the procedures feasible. An illustrative example with a real data set is also reported.