Statistical consulting is a common and vital activity undertaken by those with advanced statistical training but may not be widely available to health professionals and trainees without access to dedicated statistical consulting cores. Here, we present a novel model of statistical consulting used by a student chapter of the American Statistical Association housed in a graduate health professions university. This student‐led organization provides statistical consulting services for faculty, staff and students university‐wide. Information on the methods of advertising consulting services, the role of a professional statistician faculty advisor, resources available to consultants and university community members and the ways in which consulting services were adapted to address the barriers introduced by the COVID‐19 pandemic are addressed. Data from 108 consults performed over the past two and a half years with 88 different consultees are analysed and discussed. This article outlines an innovative model of student‐led statistical consultation for healthcare professionals and trainees and aims to provide a template for future student‐led organizations who similarly aim to perform university statistical consultations.

By now, statisticians and the broader research community are aware of the controversies surrounding traditional hypothesis testing and p values. Many alternative viewpoints and methods have been proposed, as exemplified by The American Statistician's recent special issue themed “World beyond p<0.05.” However, it seems clear that the broader scientific effort may benefit if alternatives to classical hypothesis testing are described in venues beyond the statistical literature. This paper addresses two relevant gaps in statistical practice. First, we describe three principles statisticians and their collaborators can use to publish about alternatives to classical hypothesis testing in the literature outside of statistics. Second, we describe an existing BIC‐based approximation to Bayesian model selection as a complete alternative approach to classical hypothesis testing. This approach is easy to conduct and interpret, even for analysts who do not have fully Bayesian expertise in analyzing data. Perhaps surprisingly, it does not appear that the BIC approximation has yet been described in the context of “World beyond p<0.05.” We address both gaps by describing a recent collaborative effort where we used the BIC‐based techniques to publish a paper about hypothesis testing alternatives in a high‐end biomechanics journal.

We investigate a natural Lie group structure that correlations possess, interpreted as a special case of the quotient manifold structure for correlation matrices. The one‐dimensional setting of the general theory gives rise to additional structure in the form of a natural associative multiplication of correlations making the space of correlations into a Lie group. We explicitly compute left‐invariant vector fields, exponential and logarithmic mappings, and ultimately, a closed‐form distance formula between correlations. The Lie group formalism is then applied to an application in anomalous correlation detection. The advantage of the Lie group method illustrated with time series data of stock prices.

This paper develops a novel non‐parametric test for testing the high‐dimensional alpha in linear asset pricing models, where the number of securities can be much larger than the time‐dimension of the return series. The asymptotic null distribution and the local power property are established for a class of weighted spatial‐sign tests, which results in an optimal test INST by choosing the weight function as the inverse of the norm. The INST test is optimal in the sense that it is locally most powerful within this class. As a non‐parametric test, the INST test is also robust to the departures from normality of the error distribution. Monte Carlo simulation and empirical study with real financial data show the superiority of INST test in terms of both robustness and efficiency.

Abstract A substantial number of statisticians work in isolated domain science departments without access to support networks and resources typical of larger statistical units. These isolated statisticians face many challenges including limited professional networks, non‐traditional roles with idiosyncratic expectations and unique career paths. Furthermore, “the curse of success” lies ahead, as success of the department may only lead to greater demand and lagging support for the services needed. We provide suggestions for career development in an isolated environment by urging statisticians to build their network and become engaged with other statisticians to reduce isolation and identify mentors. We discuss the importance of regular communication with their supervisors to define one's role and navigate the change in expectations in the transition from a supporting role to a co‐investigator to a core leader. To help with “the curse of success,” we explain how to develop a business model with a plan for future expansion and professional development. We conclude that isolation should not be perceived as a limitation. Opportunities and resources exist within the American Statistical Association (ASA) to reduce isolation and support career development. However, there is a need for a formal career path along with evaluation criteria and example business models to support isolated statisticians striving to become core leaders.

In recent days, a combination of finite mixture model (FMM) and hidden Markov model (HMM) is becoming popular for partitioning heterogeneous temporal data into homogeneous groups (clusters) with homogeneous time points (regimes). The regression mixtures commonly considered in this approach can also accommodate for covariates present in data. The classical fixed covariate approach, however, may not always serve as a reasonable assumption as it is incapable of accounting for the contribution of covariates in cluster formation. This paper introduces a novel approach for detecting clusters and regimes in time series data in the presence of random covariates. The computational challenges related to the proposed model has been discussed, and several simulation studies are performed. An application to United States COVID‐19 data yields meaningful clusters and regimes.

We introduce new machine learning methods for clustering traders who are actively trading in a modern electronic exchange which uses a matching engine to track aggregate and individual‐level limit order books. Each trader's individual limit order book is centered (with the current best bid and ask prices acting as a central reference), and the patterns in the individual limit order books are identified using a Wasserstein distance. The method is illustrated using simulated limit order book data and limit order book data from a stock exchange in Canada.

In clinical trials and causal inference, it is often critical to balance treatment allocation over influential covariates. In big data era, the number of covariates is usually very large, among which only a small fraction of them are influential to the response variable due to sparsity. However, existing studies assume that all influential covariates are known, fixed and given. In this article, we propose a procedure that can select the influential covariates from a diverging number of candidates and keep the allocation balanced among the important covariates simultaneously. Under mild regularity conditions, we show that the proposed procedure can pick out important covariates and balance treatment allocation among the important covariates consistently. Further, balancing treatment allocation can help the selection of important covariates, whereas picking out important covariates can make the randomization more efficient. Numerical studies support our theoretical discoveries for the proposed procedure. We also apply our method to a virtual redesign dataset of advertising vehicle choosing and show the advantages of the proposed procedure.

Understanding how a society views certain policies, politicians, and events can help shape public policy, legislation, and even a political candidate's campaign. This paper focuses on using aggregated, or interval censored, polling data to estimate the times when the public opinion shifts on the US president's job approval. The approval rate is modelled as a Poisson segmented (joinpoint) regression with the EM algorithm used to estimate the model parameters. Inference on the change points is carried out using BIC based model averaging. This approach can capture the uncertainty in both the number and location of change points. The model is applied to president Trump's job approval rating during 2020. Three primary change points are discovered and related to significant events and statements.

Cumulant coefficients are the building blocks of Edgeworth expansions and other analytic methods for statistical inference, such as bias reduction. Suppose that is a standard estimate of an unknown vector w and that t(·) is a smooth function on w. We show that is a standard estimate of t(w). We give its cumulant coefficients in terms of those of and the derivatives of t(w). These “chain rules” can then be used to get the cumulant coefficients of a smooth function of . The results are illustrated by two numerical examples.