Associate Professor Ray Watson
Consulting
I have been involved in many statistical consulting jobs — mainly for clients from other departments within the university: including Surgery, Civil Engineering, Agriculture, Otolaryngology, Anatomy, Psychology and Pathology; but also a number of external clients: including Meat Standards Australia, ACA, Esso and Telstra.
My major consulting activity for the past few years has been my ongoing involvement in developing a meat quality measure for Meat Standards Australia. Other smaller jobs I have been involved with recently included: an analysis of experimental data for zebra blood gases; a survey for the Australian Tax Office; analysis of house pay-outs for various gambling games. In addition, I am often used as the consultants' consultant.
Research
My main research interest has been in stochastic processes and in particular population models — particularly in producing usable approximations for stochastic models to overcome the inadequacy of deterministic models on the one hand and the intractability of the stochastic models on the other. This has led to an interest in martingale central limit theorems and random time scale transformations as useful tools in such endeavours.
I retain an active interest in population modelling and inference for such models. However, my research interests are quite diverse and include work in education and applied statistics. I have published papers in distribution theory, sample size determination and survival analysis.
Current research projects include:
- Estimating the mean, even in adverse circumstances (Ian Gordon).
- Binomial sequences and their asymptotic behaviour.
- Group testing (Graham Hepworth).
- Epidemic modelling and data analysis (Paul Yip).
- Inference in recapture and removal experiments (Paul Yip).
- Sport statistics and modelling.
- Modelling probability of fault detection (Cayt Harding, Caz Nickson).
- Models for discrete failure times and their statistical analysis.
Some of my current and past research papers:
In progress
- Hepworth & Watson, “Debiased estimators for group testing”.
- Harding, Watson, Sharpe & Hugo, “An improved 95% confidence limit for probability of detection using a log-normal model”.
- Yip, Watson & Lau, “Estimation of the number of people in a demonstration”
- Watson, Gee, Polkinghorne & Porter “Consumer assessment of eating quality – Development of protocols for MSA testing”.
Submitted papers
- Xi, Watson & Yip, “Estimation in capture-recapture models when covariates are subject to measurement errors or missing data”. submitted to Biometrics
- Watson “Meta-analysis of studies regarding effects of hormone growth promotant use on beef palatability measured by objective testing and by consumer studies” submitted to Australian Journal of Experimental Agriculture.
- Watson “Development of a grading model to predict consumer palatability of the carcass” submitted to Australian Journal of Experimental Agriculture.
Accepted papers
- Xi, Watson & Yip, (2007) “The minimum capture proportion for reliable estimation in capture-recapture models”, accepted Biometrics.
- Chen, Yip & Watson, (2007) “Estimation of vaccine efficacy and the vaccination threshold”, accepted Statistics in Medicine.
- Xi, Yip & Watson (2007) “A Unified Likelihood-based Approach for Estimating Population Size in Continuous-Time Capture-Recapture Experiments with Frailty” accepted Biometrics.
- Yip, Yang, Ip, Law & Watson (2007) “A study on financial debts and suicides in Hong Kong SAR”, accepted by Journal of Applied Social Psychology
Recent published papers
- Xi, Yip & Watson (2006) Estimating population size in a continuous-time removal experiment with a known sub-population size ratio. Environmental and Ecological Statistics, 13:109-126.
- Fang, Watson, Yan & Yip (2004) A procedure for complete fault detection with a removal process . J. Stat. Planning & Inference, 117: 1-14.
- Liu, Yip & Watson (2003) Removal process estimation of population size for a population with a known sex ratio. Environmental and Ecological Statistics, 10: 281-292.
- Ng & Watson (2002) A non-identifiability in modelling discrete failure times Australian & New Zealand Journal of Statistics, 44: 467-478.
- Watson & Williams (2002) Martingale generating functions for Markov chains. J. Stat. Planning & Inference. 103: 39-49.
Some older papers
- Gordon & Watson (1996) The myth of continuity-corrected sample size formulae. Biometrics. 52: 72-77.
- Watson & Yip (1993) A Bivariate counting process. J. Appl.Prob. 30: 353-364.
- Adams & Watson (1989) A discrete time parametric model for survival analysis. Aust. J. Statist. 31: 365-384.
- Watson & Gordon (1986) On quantiles of sums. Aust. J. Statist. 28: 192-199.
- Watson (1981) An application of a martingale central limit theorem to the standard epidemic model. Stoch. Proc.Applicns. 11: 79-89.
- Watson (1980) On the size distribution for some epidemic models, J. Appl. Prob., 17: 912-921.
- Anderson & Watson (1980) On the spread of a disease with gamma distributed latent and infectious periods, Biometrika, 67: 191-198.
- Watson (1975) An application of martingale methods to conflict models,
Opns. Res., 24: 380-382.
Teaching
I have been most closely involved with the second year course for Statistics majors, 620-201 (Probability) and 620-202 (Statistics), but I have taught a wide variety of undergraduate courses: indeed, I had the unique distinction of, at some time or another, having been involved in teaching all of the units offered by the now non-existent Statistics department, including third year units in Probability, Statistical Inference, Stochastic Processes, Decision Theory, Linear Models, Design of Experiments, Time Series, Distribution-free Methods, Sample Surveys and even Operations Research and Linear Programming. I have also given several fourth year courses including courses in Stochastic Processes and Survival Analysis.
In 2007, I am involved in teaching 620-270 (Applied Statistics) and 620-372 (Applied Statistical Inference).