Statistical Practice 2
HMS771
Overview
To enrol in this unit, you must be accepted into a course from the provider.
Read before you start
Level of study: What does Postgraduate mean?
Postgraduate
EFTSL: What does EFTSL mean?
0.125
Delivery Method: What does delivery method mean?
Fully Online
Prerequisites: What are the prerequisites?
Duration:
13 weeks
Government loans available:
FEE-HELP FEE-HELP
Domestic student fee:
$1,950.00 (AUD)
International student fee:
$2,175.00 (AUD)
Description
This unit provides an extension of statistical inference to testing means for more than two groups, using analysis of variance for single factor and two factor designs as well as mixed design analysis of variance. Included are: an introduction to power analysis; inference for simple regression, testing of regression assumptions using residual analysis and data transformations; non-parametric methods for testing medians in single, related and independent groups (eg. Sign, Wilcoxon, Friedman, Kruskal-Wallis); and analysis and interpretation of cross-tabulations, including measures of association. Special emphasis is placed on reporting the results.
Please note: Assessment values are indicative only; details will be advised at the start of the unit.
Enrolment Restriction
In order to enrol in this unit, you must be accepted into one of the following courses:
- Graduate Certificate of Science (Applied Statistics)
- Graduate Diploma of Science (Applied Statistics)
- Master of Science (Applied Statistics)
If you wish to seek approval to enrol in this unit without being accepted in a course, please contact OUA regarding the process.
Prerequisites
Recommended prerequisites
You are recommended to have completed the following unit(s) or have equivalent knowledge before starting this unit:
- HMS770 — Statistical Practice 1
You should complete HMS770 before studying this unit. If you have some statistical background it is possible to complete HMS770 concurrently with HMS771.
Special Requirements
- Additional materials
- Other special requirement — You will require access to a recent version of SPSS - at least the SPSS Grad Pak.
Assessment
- Assignment 1 (12%)
- Assignment 2 (20%)
- Invigilated Exam (60%)
- Quiz — Online Quizzes (8%)
Learning Outcomes
Analysis of Variance
At the completion of this unit students will be able to:
- understand when it is appropriate to use the Analysis of Variance
- perform and interpret an analysis of variance for the single factor independent groups design
- perform and interpret an analysis of variance for the single factor within subjects design
- perform and interpret a completely randomised, factorial analysis of variance
- perform and interpret a mixed design, factorial analysis of variance
- perform power analysis for analysis of variance to determine the number of subjects required to achieve a given power
- perform a power analysis for analysis of variance to establish if a study was sufficiently powerful when the F-ratio is not significant
- test the assumption underlying analysis of variance
- write reports on analysis of variance
- write a description of the relationship between two metric variables, based on the graphs and statistics produced using SPSS
- perform a significance test, both on Pearson's r and on the slope of the regression line
- perform a power analysis for correlation.
Bivariate Regression and Transformation
At the completion of this unit students will be able to:
- understand the theory behind simple linear regression
- use SPSS to construct residual plots, histogram and QQ plot of standardised residuals and hence determine whether or not the assumption of normality of errors has been met
- from a residual plot to determine whether or not the assumptions of linearity and homogeneity of variance have been met and if there are any possible outliers
- recognise when a transformation of the data would be appropriate
- use the ladder of transformations to systematically investigate different transformations
- use the coefficient of determination and residual plot to help you decide which transformation results in the best model for the data
- use SPSS to transform data and investigate the results.
Non-parametric tests
At the completion of this unit students will be able to:
- determine which nonparametric test is applicable for a single sample or a pair of related samples or two or more independent groups, given the level of measurement of the data
- use SPSS to carry out this test
- determine whether of not a nonparametric test or a parametric test is more appropriate
- describe the relationship between two categorical variables by appropriately percentaging a table
- understand and interpret chi-square test and Fisher's exact test
- understand what is meant by a PRE measure of association
- calculate lambda from first principals
- use SPSS to calculate other measures of association
- use an appropriate measure of association to classify the strength of a relationship as weak, moderate or strong.
Topics
This unit addresses the following topics.
| Number | Topic |
|---|---|
| 1 | ANOVA |
| 2 | Review of variance and t-tests |
| 3 | Introduction to the analysis of variance - the singel factor, independent and groups design |
| 4 | Using SPSS to produce an analysis of variance |
| 5 | Reporting an analysis of variance |
| 6 | Analytical comparisons in the single factor independent groups design |
| 7 | The completely randomised factorial design and report writing |
| 8 | Analysis of variance for the single factor within subjects and report writing |
| 9 | The mixed factorial design |
| 10 | Regression and nonparametric methods |
| 11 | Correlation, regression and inference for regression |
| 12 | Investigating the assumptions of regression |
| 13 | Data transformations |
| 14 | Non-parametric methods for single and related groups |
| 15 | Association between two ordinal variables |
| 16 | Association between two categorical variables |
| 17 | Review of the chi-square test |
| 18 | Measuring the strength of the association in tables |
| 19 | Non-parametric methods for single and related groups |
| 20 | Non-parametric methods for independent groups |
| 21 | Choosing the correct statistical test |
Study Resources
This unit is delivered using the following methods and materials:
Instructional Methods
- Discussion Forum/Discussion Board
- Online Quizzes/Tests
- Online assignment submission
- Web links
Print based materials
- Welcome Letter
Online materials
- Printable format materials
Relevant Courses
This unit is a core requirement in the following courses:
- Graduate Certificate of Science (Applied Statistics), Swinburne University of Tnlgy
- Graduate Diploma of Science (Applied Statistics), Swinburne University of Tnlgy
- Master of Science (Applied Statistics), Swinburne University of Tnlgy
This unit may be eligible for credit towards other courses:
- Many undergraduate courses on offer through OUA include 'open elective' where any OUA unit can be credited to the course. You need to check the Award Requirements on the course page for the number of allowed open electives and any level limitations.
- In other cases, the content of this unit might be relevant to a course on offer through OUA or elsewhere. In order to receive credit for this unit in the course you will need to supply the provider institution with a copy of the Unit Profile in the approved format, which you can download here. Note that the Unit Profile is set at the start of the year, and if textbooks change this may not match the Unibooks textbook list.
