The Hausdorff Space and Applied Statistics: A View from the U.S.S.R.

The article focuses on the debate in the pages of the Soviet statistical journal Vestnik Statistiki as to the existence and relevance of applied statistics as a separate scientific discipline. The contents of four letters to the editor and relevant editorial comments that appeared in this journal between October 1985 and July 1987 are analyzed.     

Knot selection for least-squares and penalized splines

Two new stochastic search methods are proposed for optimizing the knot locations and/or smoothing parameters for least-squares or penalized splines. One of the methods is a golden-section-augmented blind search, while the other is a continuous genetic algorithm. Monte Carlo experiments indicate that the algorithms are very successful at producing knot locations and/or smoothing parameters that are near optimal in a squared error sense. Both algorithms are amenable to parallelization and have been implemented in OpenMP and MPI. An adjusted GCV criterion is also considered for selecting both the number and location of knots. The method performed well relative to MARS in a small empirical comparison.     

Simulation study of wildlife density estimators

Estimators of the mean individual calling rate and density of a wildlife population in a fixed area are derived on the basis of aural information ob-tained by a listener within the fixed area. An estimator of density assuming known calling probabilities and independent callers is minimum variance un-biased. Estimators of density based on unknown calling probabilities are derived under the assumption of independent callers and a Poisson distribu-tion of calls for each caller. The effectiveness and robustness of the estimators are investigated by simulation     

Critical reflection workshops and knowledge exchange: findings from a Scottish project

In 2013, academics from a Scottish university came together with social work managers and practitioners from two local authorities (LAs) in Scotland to run a knowledge exchange (KE) project co‐sponsored by the Economic and Social Research Council and the LAs. The project's aim was an ambitious one: to contribute to culture change in the children and families' departments in the two partner agencies. The project grew out of an earlier KE venture that had explored ways of engaging better with involuntary service users in social work; it thus both anticipated and reflected wider concerns about services for children that were also demonstrated in Munro's Review of Child Protection. The KE project had three components: training for managers, practitioner research projects and critical reflection workshops. Whether, and to what degree, the KE project changed culture is not the focus of this paper, which is written jointly by academic researchers and practitioners. Instead, one element of the KE project, namely the critical reflection workshops, is discussed. Findings provide strong evidence of the pressures currently experienced by children and families' services in the UK public sector. They also indicate how important good relationships are in building meaningful KE.     

Motivation and mission in the public sector: evidence from the World Values Survey

It is well-recognised that workers may have intrinsic—as well as extrinsic—motivations. Previous studies have identified that public sector workers typically have a higher level of intrinsic motivation, compared to workers in the private sector. This paper compares (measures of) intrinsic motivation among 30,000+ workers in the two sectors across 51 countries using data from the World Values Survey. We find that public sector workers exhibit higher intrinsic motivation in many countries, but that this is not a universal relationship. One possibility is that public sector mission may influence whether or not motivated workers choose to work in the sector. In support of this, we show that the level of (public) corruption—which plausibly affects mission—can explain some of the variation across countries in the proportion of motivated workers in the sector.     

Utility values in health technology assessments: a statistician's perspective

This paper provides an introduction to utilities for statisticians working mainly in clinical research who have not had experience of health technology assessment work. Utility is the numeric valuation applied to a health state based on the preference of being in that state relative to perfect health. Utilities are often combined with survival data in health economic modelling to obtain quality‐adjusted life years. There are several methods available for deriving the preference weights and the health states to which they are applied, and combining them to estimate utilities, and the clinical statistician has valuable skills that can be applied in ensuring the robustness of the trial design, data collection and analyses to obtain and handle this data. In addition to raising awareness of the subject and providing source references, the paper outlines the concepts and approaches around utilities using examples, discusses some of the key issues, and proposes areas where statisticians can collaborate with health economic colleagues to improve the quality of this important element of health technology assessment. Copyright © 2014 John Wiley & Sons, Ltd.     

A nonparametric test for bivariate circular symmetry based on the empirical cdf

A nonparametric test for circular symmetry about 0 in a continuous bivariate distribution is proposed. The test is of the von Mises type, based on the empirical cdf of the sample, expressed in polar co-ordinates. However, the test is independent of the choice of the polar axis. The asymptotic form of the test statistic is obtained by considering the weak convergence of the empirical process to a limiting Gaussian process. The asymptotic distribution of the test statistic is found explicitly, both under the null hypothesis and under simple alternatives. The test is shown to be consistent against all alternatives.     

Interruptible supersaturated two-level designs

Interruptible designs possess a robustness against possible premature termination of an experiment. We consider such two-level designs for a first-order model and present interruptible sequences which lead to the D-optimal saturated design for four to nine factors if not interrupted. Premature termination of the experiment at any stage results in a supersaturated design with minimum loss of information about the factors. The loss for these designs, which is measured by the pairwise orthogonality between columns, is compared with that of the worst case f o r randomly ordered sequences.     

Examination of a parsimonious factor screening technique:screening K = 2N - 1- 1 factors in n runs

In many experimental situations (particularly in computer simulation studies) a large number of potentially important factors exist. Because of time and budget limitations, it is imperative to screen these factors in order to identify a subset that should be subjected to more detailed' examination. This paper examines the performance of a factor screening technique when it is known that there is at most one active factor (i.e., a factor which has an effect on the response of interest). Although the technique works perfectly in the deterministic case, care must be taken when random error is present.