Statistics Short Courses

Statistical Methods for Research Workers - 2016

In many disciplines, researchers wishing to publish are asked to provide a rigorous statistical analysis. Reviewers are often specific about what statistical measures they want included. Why wasn't multiple logistic regression used? Was an appropriate sample size determined a priori?

Statistical analyses require specialised software to perform calculations. In this course we use the free statistical program R, although researchers may have another statistical package available to them.

How does one decide which statistical procedure is the most appropriate? What do all the pages of the printout mean?

This course is designed as an overview of statistical design and analysis for researchers. There is emphasis on understanding the concepts of statistical procedures (with a minimum of mathematics, although some will be discussed) and on interpreting computer output. This course is designed to help you, the researcher. It assumes you did an undergraduate statistics subject, and you may need a refresher.

Instructions on how to obtain computer printouts will be provided with an emphasis on interpreting the computer printout (most packages produce similar printouts). There will be computing lab sessions throughout the course.

Do you need to have previous experience using the program R?

Our statistics short course does not require previous knowledge of R; we spend a fair amount of time introducing and using R Studies (an R IDE). However, it is helpful if you have previous experience with statistical packages like SAS, SPSS or STATA. It is also helpful if you have some basic programming skills.

Please note that R and R Studio are free to use and are both available for Windows/Mac/Linx platforms. The links are provided below:

R Project

R Studio

You can play around with the software before deciding on the course. Here is a really good website with some introductory notes for doing statistics in R.

Course Outline

Introduction

Types of experiments, scales of measurement, which method to use.

Software

In this course the statistical software package used is R we do not use or demonstrate SPSS.

Summarising and Graphing Data

Ways of presenting data (histograms, boxplots), measures of centre and spread, analysing tables, correlation, and confidence intervals.

Comparing Groups

Hypothesis testing concepts-power, significance, P-value. Comparing two groups (t-tests, Wilcoxon). Comparing many groups - ANOVA or Kruskal-Wallis - multiple comparison tests, required sample size and repeated measures.

Finding Relationships

Correlation, predicting relationships (regression - simple).

Other topics covered:

  • Survival analysis

  • Linear mixed models for longitudinal and clustered data

  • Generalised linear models (specifically logistic and Poisson regression)

Duration: 3 days, 9.30am to 5.00pm daily

Dates: 28, 29 and 30 June 2016

Details and registration: Please visit the Stats Central website

Contact:

Stats.central@unsw.edu.au