STA2216 - Data analysis for science - 2017

This unit is designed to develop an understanding of some of the most widely used methods of statistical data analysis, from the view point of the user, with an emphasis on planned experiments. Students will become familiar with at least one standard statistical package. Topics covered include: parametric and nonparametric procedures to compare two independent and matched samples; review of simple linear regression; multiple linear regression - analysis of residuals, choice of explanatory variables; model selection and validation; nonlinear relationships; introduction to logistic regression; basic principles of experimental design; one-way and two-way analysis of variance models; planned and multiple comparison techniques; power and sample size considerations in design; usage of some available statistical packages including Minitab and/or SPSS, data preparation, interpretation of output.

On completion of this unit students will be able to:

1. Recognise the requirements for design of an effective experiment and the nature of data arising from these situations;
2. Demonstrate an understanding of some of the important parametric and non-parametric methods of statistical data analysis, including analysis of variance, multiple linear regression and logistic regression;
3. Identify and apply an appropriate statistical technique for analysing a given design/ data set;
4. Formulate a model relating a response variable to a number of given independent variables;
5. Use a statistical package for applying statistical techniques covered in the unit.

Assignments (3): 30%

Mini-project report: 10%

Examination (2 hours): 60%

Three 1-hour lectures and one 1-hour workshop per week

See also Unit timetable information

Dr Ito Wasito

Tropical environmental biology

STA1010 or SCI1020

MTH2232, STA2032