Te Tari Pāngarau me te Tatauranga
Department of Mathematics & Statistics

STAT501 Statistical Modelling for Research

First Semester
18 points
Not available in 2018

The aim of this paper is to provide postgraduate students with many of the important statistical tools that they require in their research. Students will gain experience in using modern statistical software (R and WinBugs).

Paper details

We cover the basics of probability through to the fitting of complex models. There will be an emphasis on the practical analysis of real data.

Potential students

Postgraduate research students (outside the Department of Mathematics and Statistics) who have taken at least a first-course in statistics and want to become familiar with modern methods of analysis and software.

Main topics

  • Probability and Distributions
  • Fitting Models to Data
  • Normal linear models
  • Generalised linear models
  • Model-selection
  • Model-checking
  • Data-collection


Ideally STAT 110/115 or equivalent. Enrolment in a research-based postgraduate programme.

Required text


Some references

Online access to all of these is available via the University Library

  • Gelman and Hill Data Analysis Using Regression and Hierarchical Models
  • Maindonald and Braun Data Analysis and Graphics Using R
  • Zuur, Ieno and Meesters A Beginner’s Guide to R


David Fletcher, room 219


Thursday 2-5pm, starting in Room 241 and finishing in B21 (computer laboratory), in the Department of Mathematics and Statistics.

Internal Assessment

There will be four assignments, each contributing 5% towards the fi nal mark, and a project worth 50%.

Exam format

2-hour exam

Final mark

While we strive to keep details as accurate and up-to-date as possible, information given here should be regarded as provisional. Individual lecturers will confirm teaching and assessment methods.
These graphs show data from a study of New Zealand sea lion pups at three breeding colonies on the Auckland Islands. The points are the estimated number of pups that day (with standard error bars). A fourth-order polynomial was fitted to the data at each site in order to estimate the date of peak pup-production. Parametric bootstrapping was used to provide an estimate of the uncertainty in these dates. The “peak-dates” were 13th, 14th and 3rd January for Sandy Bay, Dundas Island and Pebble Point respectively (with standard errors of 0.5, 1.8 and 0.8). We could probably come up with similar estimates of the dates “by eye” but providing an idea of their uncertainty would be much harder.