# ANALYSES OF LINEAR MODELS

STA4LM

2018

Credit points: 15

## Subject outline

Modern research often involves the analysis of data for more than one variable and in this regard, linear models are the most widely used class of models. Linear models relate a response variable to one or more explanatory variables enabling researchers to answer important research questions and make predictions about how variables will respond. These methods are used in many areas including biological science, economics, engineering, medical science and psychological science. Topics covered in this subject include simple and multiple linear regression, response and explanatory variable transformations, ANOVA and ANCOVA, as well as more modern methodologies such as generalized linear models and linear mixed effects models. This subject has a stong emphasis on preparing students for future careers in statistics. This subject is co-offered with STA3LM although assumes a deeper level of understanding of the subject content.

SchoolSchool Engineering&Mathematical Sciences

Credit points15

Subject Co-ordinatorAgus Salim

Available to Study Abroad StudentsYes

Subject year levelYear Level 4 - UG/Hons/1st Yr PG

Exchange StudentsYes

## Subject particulars

### Subject rules

Prerequisites Must pass one of the following: STA2ASM, STA2ABS, STA2AMS, STA2MD, STM2PM or approval by Honours or Masters by Coursework Coordinator

Co-requisitesN/A

Incompatible subjects STA3LM

Equivalent subjectsN/A

Special conditionsN/A

Resource TypeTitleResource RequirementAuthor and YearPublisher
ReadingsIntroduction to Linear Regression AnalysisRecommendedMontgomery, DC, Peck, EA and Vining, G 2006WILEY, 4TH EDITION.

## Graduate capabilities & intended learning outcomes

01. Present clear, well structured proofs of important fundamental linear model results that include appropriate use of statistical and mathematical vocabulary and notation.

Activities:
Weekly problem classes involve theoretical derivations of results introduced in lectures. 2 assignments consist of up to 50% assessed theoretical derivations.

02. Describe and use key analytical linear modelling tools including a justification of appropriate usage based on known model/data conditions.

Activities:
Appropriate usage of methodologies is discussed and modelled via example in lectures. Weekly computer lab classes involve interpreting, including a justification of, computer output relating to analyses of real data sets. 2 assignments consist of up to 50% assessed usage and justification of methodolgies.

03. Implement and document various strategies to identify and account for model inadequacies.

Activities:
Lectures in week 3 onwards introduce students to ways in which model conditions can be critiqued and to the manner in which inadequacies can be corrected. Computer lab classes from week 3 onwards and assignments 2 and 3 require students to defend the use of methodolgies by implenting model checking as well as requiring the correction of inadequacies when necessary.

04. Present written and oral communications of statistical results clearly in a manner that can be understood by expert and lay audiences.

Activities:
Simple written summaries based on real data analyses are modelled weekly in lectures. Weekly computer lab classes in part involve students writing simple evidence based conclusions. 3 assignments also partly require students to prepare such simple conclusions. 25% of the total assessment is allocated to a consulting role-roleplay which requires students to work in teams to produce a statistical analysis based on real data. Thie role-play culminates in each student individually verbally communicate results of the analysis to the lecturer where the lecturer acts as a scientist with little statistics background.

05. Work efficiently and effectively as a member of a team to produce a statistical analyses based on real data.

Activities:
25% of the total assessment is allocated to a consulting role-roleplay which requires students to work in teams to produce a statistical analysis based on real data.

06. Justify the choice of an appropriate error term correlation structures for linear mixed effects model analyses.

Activities:
Questions specific to STA4LM students in the week 12 computer lab and practice class will involve the using the internet to explore various correlations structure. The student is expected decribe the most commonly used of these. The final exam will also require justification of correlation structure.

## Subject options

Select to view your study options…

Start date between: and    Key dates

## Melbourne, 2018, Semester 2, Day

### Overview

Online enrolmentYes

Maximum enrolment sizeN/A

Enrolment information

Subject Instance Co-ordinatorAgus Salim

### Class requirements

Computer Laboratory Week: 32 - 43
One 1.0 hours computer laboratory per week on weekdays during the day from week 32 to week 43 and delivered via face-to-face.

Lecture Week: 31 - 43
Two 1.0 hours lecture per week on weekdays during the day from week 31 to week 43 and delivered via face-to-face.

Practical Week: 32 - 43
One 1.0 hours practical per week on weekdays during the day from week 32 to week 43 and delivered via face-to-face.