This unit will provide students with an understanding of experimental design and data analysis using the R programming language. This will be achieved through supported IT practicals, self-study practicals and independent reading. Topics will include: principles of experimental design; principles of frequentist statistics, analysis of variance, regression, analysis of categorical data and simple non-parametric statistics.
Aims and Objectives
This unit aims to equip candidates with the core skills needed to design experiments and analyse observational and experimental data. They will learn about: (1) principles of experimental design and sampling; (2) basic principles of frequentist interference; (3) the link between resampling methods and frequentist ideas; (4) how to choose an appropriate statistical model for your data (t-tests, ANOVA and regression); (5) model fitting, evaluation and inference using the R statistical programming language. The unit will focus on practical, rather than theoretical treatments of data analysis, using simulations in R to motivate learning. The main topics are:
Understanding statistics. You will learn how to: (1) Explain frequentist concepts such as population vs. samples, null hypotheses and p-values; (2) Apply bootstrap and permutation tests to evaluate differences between two means; (3) Use a t-test to evaluate the significance of mean differences; (4) Construct confidence intervals for sample statistics.
Principles of Experimental Design. You will learn how to: (1) Explain the three fundamental principles of experimental design (replication, randomisation and blocking); (2) Evaluate the strengths and weaknesses of different sampling protocols; (3) Design experiments using factorial and blocked structures; (4) Demonstrate an awareness of more complex designs.
Analysis of variance. You will learn how to: (1) Use geometric reasoning to explain why differences in variances and F-tests are used to evaluate significance in ANOVA; (2) State the assumptions of ANOVA; (3) Specify and fit one-way and two-way ANOVA in R; (4) Interpret ANOVA tables and summaries of fitted coefficients; (5) Report the results of ANOVA in tables and graphs.
Regression. You will learn how to: (1) Use geometric reasoning to explain how a "line of best fit" is found in simple regression; (2) State the assumptions of regression; (3) Interpret summaries of fitted coefficients; (4) Report the results of simple regression in tables and graphs.
Practical Model Fitting. You will learn how to: (1) Use R to check model assumptions and critically evaluate model fit; (2) Choose appropriate transformations to remedy problems with a fitted model.
Delivery Method: 9 Supported IT Practicals, 9 Self-study IT Practicals, and Independent Study
Student contact hours: 18
Assessment method: 50% Examination and 50% Coursework
Feedback: Students will receive feedback during the course.
Please go to MOLE for more information on APS240