| Unit name | Understanding Data: Experimental Design and Statistics for Life Scientists |
|---|---|
| Unit code | BIOLM0006 |
| Credit points | 20 |
| Level of study | M/7 |
| Teaching block(s) |
Teaching Block 1 (weeks 1 - 12) |
| Unit director | Professor. Cuthill |
| Open unit status | Not open |
| Pre-requisites |
There are no pre-requitites however it helps if you have done some sort of undergraduate course in statistics. That is, you know about means, standard deviations, variances, standard errors, and confidence intervals. You know about chi-squared tests, some non-parametric tests, t-tests, one-way ANOVA, correlation and simple regression. When we say you “know” these things, we mean you have covered these items before, remember more-or-less what they were about, or if not, have some notes or a book that you could look them up in. If you haven't done such a course, you will still be able to do the unit as long as you have the desire to understand data. In advance of the unit, we will make available a self-assessment tool with guidance towards appropriate support materials so that you can be suitably prepared for the course. |
| Co-requisites |
none |
| School/department | School of Biological Sciences |
| Faculty | Faculty of Life Sciences |
Being able to frame a research question in an appropriate manner, to gather the right data and analyse such data correctly, are key skills in the Life Sciences. Perhaps THE key skills for any empirical scientist.
This unit is about designing research and data collection programmes, and the more advanced statistics that most working biologists (and, indeed, medical researchers, psychologists, geographers, and economists) need to know about. In the first lectures we will cover all the core principles, to clear out some of the cobwebs from half-remembered undergraduate courses and, indeed, dispel the many common misconceptions. A key aim of the course is that you come away understanding enough of the theory (and practice) to be able talk sensibly to REAL statisticians – to describe your problems in ways they can understand and to understand their proposed solutions. This is a good goal for you to aim for – not to BE a statistician but to be able to have an informed discussion or collaboration with them if your data analysis problems go beyond the basics. That said, what we will teach you as ‘the basics’ are fairly advanced – you’ll be able to tackle 99% of the statistical problems most biologists (and medical researchers, psychologists, geographers, etc.) face.
Topics include include experimental design, graphical display and data exploration tools, basic univariate parametric and non-parametric tests, Monte Carlo and computer-intensive methods, multi-factor ANOVA, repeated-measures ANOVA, multiple regression, polynomial regression, General (and Generalized) Linear Models, Mixed and Multi-Level Models, and multivariate methods; the unit ends with an introduction to Bayesian approaches to statistics. These will all be taught from scratch, backed up by hands-on practical sessions and problem-solving exercises. The unit will also provide an introduction to the computer software packages SPSS and R, the former probably the most widely used package in academia and industry, the latter the most flexible, portable and powerful.
An appreciation of experimental design coupled with sampling methods for data collection.
An understanding of the background to statistical methods and how to use them to analyse data, particularity how to apply the appropriate methods for the data.
An understanding of how to display and interpret results both graphically and in writing.
Seminars, practicals and workshops
Practical classes and problem-solving exercises (30%)
Short answer/multiple choice tests (30%)
End-of-unit, open book, data analysis and presentation exercise (40%)
Extensive handouts will provided as part of the unit, but useful reference texts include:
Quinn, G.P. & Keough, M.J. (2002) Experimental Design and Data Analysis for Biologists. Cambridge: CUP.
Logan, M. 2010. Biostatistical Design and Analysis Using R. Chichester: Wiley-Blackwell.
Dalgaard, P. 2002. Introductory Statistics with R. New York: Springer.
