Design and Power Considerations for Cluster-Randomized Cross-over Designs (CRXO) with Bounded Discrete Endpoints

Thursday, 27 June 2013
9:30am - 10:30am
Room 515, Level 5, Melbourne School of Population and Global Health, Melbourne University
207 Bouverie Street
Carlton 3053

CRXO designs are not uncommon in public health research and bounded discrete endpoints, such as pain scales, are common.  The justification of sample size (i.e. the number of clusters, periods and individuals) is generally based on power calculations that assume continuous-scale, normally-distributed, primary endpoints.  In this talk, various approaches are reviewed and some simulation-based results are reported for the power of a conventional test procedure when it is “misapplied” to a bounded discrete endpoint.


Assoc. Prof. John Reynolds

Head, Biostatistics Unit, Faculty of Health
Deakin University

John Reynolds joined the Faculty of Health at Deakin University in July last year as the Head of the new Biostatistics Unit.  John has experience in the design, management and statistical analysis of clinical trials in oncology and experiments in the food and agricultural sciences.

He previously held the position of Expert Statistician with Novartis Oncology, Basel, Switzerland and was the Lead Statistician for two new compounds (all indications) and two marketed compounds (in Chronic Myeloid Leukaemia).  He was responsible for the quality of the work of a small team of biostatisticians and programmers located in Hyderabad and Paris, and liaised with clinical investigators located in Europe and the USA.

Prior to working for Novartis, John was a Senior Statistician in, and Deputy Director of, the Centre for Biostatistics and Clinical Trials at the Peter MacCallum Cancer Centre (Peter Mac) in East Melbourne.  As well as supporting research at Peter Mac, he was the Biostatistician for the Australasian Leukaemia and Lymphoma Group (ALLG). 

Before moving into oncology biostatistics, John was the Chief Biometrician in the Department of Natural Resources and Environment (NRE), Victoria.  This position involved the professional supervision of biometricians as well as statistical consulting (design and analysis of field trials, laboratory experiments and sensory evaluation panels).  As part of his role with NRE he also contracted a portion of his time to Food Science Australia (FSA), now known as CSIRO Food and Nutritional Sciences, Werribee.

In his career to date, he has also given short courses on statistics at universities in Australia, New Zealand and Thailand, worked as a private statistical consultant, and appeared as an expert statistical witness in a product defamation case (the “Maxicrop” case) in the High Court of New Zealand.

He spent part of his early career at North Carolina State University as a graduate research assistant and post‑doctoral fellow in the Department of Statistics investigating genetic distances and the estimation of divergence times between populations. 

He completed his undergraduate training in mathematical statistics at Massey University in Palmerston North, New Zealand.