Statistical Methods in Analytical Science Workshop
9th - 10th December
Outline of the course:
- INTRODUCTION – the need for statistics; data and measurement in AS; uncertainty in sampling and measurement
- ERRORS – random and systematic error; error propagation
- SUMMARISING DATA – measures of location and spread; graphical representation of one- and two-variable sets
- DISTRIBUTIONS – the Normal distribution (in particular) and why it is important; the lognormal, Poisson and binomial distributions
- HYPOTHESIS TESTING – t-tests, F-tests and the chi-squared test
- QUALITY CONTROL – Shewhart and Cusum charts; proficiency testing and collaborative trials
- CALIBRATION & REGRESSION
- EXPERIMENTAL DESIGN – The basics of design; factorial designs and response surfaces
- SOME MULTIVARIATE METHODS – principal components analysis; cluster analysis & partial least squares
It is possible that we will not be able to cover all these topics, but hand-outs and further resources will be available.
The course will assume a basic understanding of mathematics – there will be some (simple) mathematical formulae – but will start from first principles, describing the nature of variation and how we cope with it using statistical methods. We will then discuss some of the basic methods necessary for exploratory data analysis in analytical science including the nature and structure of error, and methods for controlling it, as well as important topics such as calibration. The afternoon session will take the form of a practical workshop in which participants will design, build and fly paper helicopters.
*Research council funded students will be able to claim for their workshop, accommodation and subsistence costs following the workshop. If you are paying for your workshop by cheque, this will not be cashed unless you do not attend the workshop.