< Back

Chris Abbess

Born in 1937 with schooling in Chandlers Ford, Eastleigh and Winchester I had problems with the eleven plus.
I left school and was ‘apprenticed’ to Folland Aircraft Ltd as an aeronautical engineer with a link to Southampton University. I graduated in 1960 and worked for Follands until the design office closed down. I retrained as a maths teacher with scripture in 1962 having just been married to Yvonne.

I found work at the Sir George Monoux Grammar School in Walthamstow. This period from 1963 to 1969 was the busiest for me as we also became parents to three of our daughters. After the school reorganised I moved to what became Middlesex Polytechnic/University, primarily to teach computing, and I retrained again at City University to gain a masters in Statistics and OR. I taught Statistics and computing at most levels, and notably to the masters in road safety engineering course, which was never quite profit making, but drew in mature students from local authourites. I also joined a research group in road transport studies and became involved in road safety, regression to the mean and congestion modelling.

After retiring in 2000, teaching was the exception rather than the rule, though I do miss the contact. I have espoused the alternative transport option by volunteering as a path ranger in the National Cycle network promoted by the Sustrans Charity. I also represent my church at deanery and diocesan synods where my concern is encouraging the church to care for the whole of the environment.

Are there limits to science? A Bayesian response

Birmingham 2016

This is not a discussion about the existence of categories that are not amenable to scientific analysis but it concerns a practical approach which will attempt to acquire knowledge or wisdom in the first instance and perhaps later decide that the expenditure or effort are not worth the value of the knowledge found. In particular I wish to demonstrate the operation of the Bayesian paradigm in a simple form as having a clear methodology, is open to inspection and been shown to have considerable success in diverse areas. It starts with a phenomenon with categories of explanation termed Hypotheses, distinct and having a degree of truth (probability) assigned to each. Next come likelihoods: these are probabilities too but express the chance of the observed data being discovered in the light of the truth of the hypothesis. By assembling the joint probabilities of hypothesis and likelihoods adding them and rescaling them we get an updated view of the probabilities of the hypotheses that we started with. Moreover as more data arrives the process of updating continues. The general expectation is that the uncertainty remaining will concentrate around those hypotheses that are best in explanatory terms. That in a nutshell is what the Bayesian paradigm is about. To illustrate the working of the procedure I propose to apply it to a hoary old chestnut known as the Monty Hall problem. Details will be given on a separate sheet – there will be the need of a little arithmetic and not much else. What are the benefits of this type of approach? We need to be humble and admit to the possibility of being wrong. W need to be broad minded to include possibilities that are easily discarded. We need to be able to observe the convergence of ideas in a context that we are ready to be driven by the data rather than push our own preferences.