unitinfo

It's UWAweek 47

unitinfo

This page provides helpful information about many coursework units offered by Computer Science and Software Engineering in 2023.

The information here is not official - for official information please see the current UWA Handbook. Instead, it will help students to prepare for their future units, before the beginning of each semester, and before they have access to UWA's Learning Management System (LMS).

About the unit STAT2401 Analysis of Experiments (1st semester 2023)

View the UWA Handbook entry (includes unit outcomes and prerequisites)

Unit description:

This unit introduces the principles of randomised experiments and practical quantitative tools for analysing the data from such experiments. The emphasis is on choosing and applying appropriate methods, rather than technical details and formal inference. Topics include controlled experiments versus observational studies; different experimental designs—confounding, misinterpretation; biases—blinding and randomisation; and analysis—linear regression, linear models and analysis of variance.

Unit outcomes:

Students are able to (1) demonstrate skills in the statistical analysis of data from designed experiments and observational studies; (2) apply the fundamentals of designing random experiments in different situations; (3) articulate the use of linear models; (4) develop and apply linear models for data from real-world experiments and studies; and (5) proficiently use a statistical computer package for linear modelling.

Unit coordinator:

Dr John W. Lau
[email protected]

Unit homepage:

UWA's Learning Management System, https://handbooks.uwa.edu.au/unitdetails?code=stat2401

Unit is offered in these majors and courses:

Indicative weekly topics:

week 1	Review: Random variables, Distributions, Normal distribution, (Student's) T distribution, two sample t test, and some matrix algebras. Introduction of the software package: R and data visualization
week 2	Simple Linear Regression I: Basic Introduction of simple linear regression. Estimation of regression coefficients. Use of the lm R command
week 3	Simple Linear Regression II: Test for significance of the regression coefficients and confidence intervals for the regression coefficients. Variability partitioning, decomposing the sum of squares, and test for the significance of regression. ANOVA table. Understand the lm R output
week 4	Simple Linear Regression III: Use of the R^2, correlation coefficient. Confidence interval for the regression line and prediction interval for a new observation
week 5	Multiple Linear Regression I: Modeling response variables using multiple predictors. Regression coefficient estimates. Use of the lm R command. Regression coefficient estimator via matrix notations
week 6	Multiple Linear Regression II: Inference for multiple linear regression model. Test for significance of partial regression coefficients. ANOVA table. Understand the lm R output. R^2 and Adjusted R^2. Confidence interval for the regression line and prediction interval for a new observation
week 7	ANOVA: Including only Categorical covariates in the regression model. Introduction of dummy variables. Connection to two sample t-test. One way ANOVA model
week 8	ANOVA & ANCOVA: Design of experiment. Two way ANOVA additive model and Full Two way ANOVA with interaction terms. ANCOVA: Including both Categorical and numeric covariates in the regression model
week 9	Polynomial regression: Basic idea of polynomial regression. Coefficient estimation. Application. Prediction and overfitting
week 10	Model Diagnostics I: Model Assumptions: normality, constant variance, independence assumptions. Accessing residual plots, normal probability plots. Outlier and high leverage detection. Cook's distance
week 11	Model Diagnostics II: Multicollinearity, idea of variable selection. Transformation the data
week 12	Variable Selection: Forward/Backward model selection. Use of criteria: Akaike information criterion (AIC), Bayesian information criterion (BIC), and adjusted R^2, F-statistic/p-value

Indicative assessment:

2 Tests, 2 Assignments, and a Final Exam

Useful prior experience and background knowledge:

STAT1400 Statistics for Science, or STAT1520 Economic and Business Statistics

Useful prior programming and software experience:

R and RStudio

Hardware required for this unit:

Students are able to undertake their laboratory exercises and projects in laboratories in the CSSE building, but most students also complete work on their own laptops.
The following hardware is required to successfully complete this unit:

Standard laptop

Operating system(s) used in this unit:

Different units will use different operating systems for their teaching - for in-class examples, laboratory exercises, and programming projects.
If an operating system is REQUIRED, it will be used when marking assessments.

ANY reasonable platform

This information last updated 6:49pm Thu 20th Apr 2023

ALL units level-1 units level-2 units level-3 units level-4 units level-5 units our unit coordinators switch years 2023 2024	login chapter0 unitinfo research projects csforum csse-feedback csbreakdown cssubmit csmarks 2nd semester help fora Algebra Advanced Algorithms Cloud Computing High Performance Computing Intelligent Agents Professional Computing Relational Database Management Systems Systems Programming