64635 STATISTICAL MODELS

Year	No.	Offer	Mode	Description			Cred. Pts
97	64635 	S2  	X 	STATISTICAL MODELS        	1.00

Contents

• Staffing
• Pre-requisite(s)
• Co-requisite(s)
• Rationale
• Synopsis
• Objectives
• Topics
• Texts
• Recommended References
• Student Workload
• Assessment Details
• Other Requirements

STAFFING:

PRE-REQUISITE(S)

75612/64612 + 75626/64626

CO-REQUISITE(S)

75631/64631

RATIONALE:

Linear Models are very widely used statistical tools.  This unit gives the
student an introduction to the theory of linear models and their applications.
An appropriate computer package is used to give students practice at handling
many data sets.  There are many situations where the usual model assumptions
are not satisfied and special techniques are required and this unit also
introduces some of these techniques.

SYNOPSIS:

This unit introduces the student to statistical linear modelling utilising
an appropriate computer package.  The topics include linear statistical
models, fitting the full rank model, inferential procedures, residuals and
outliers, multicollinearity, autocorrelation, generalised linear models,
and analysis of categorical data.

OBJECTIVES:

On successful completion of this unit students will be able
to:

1. construct an appropriate linear model for a data set;
   
2. use an appropriate computer package to fit the model;
   
3. test the constructed model for goodness of fit and
   significance of parameters;
   
4. examine the residuals from the model and interpret residual
   plots;
   
5. recognise the existence of possible outliers and how to deal
   with them, detect and deal with outliers;
   
6. understand the concepts and procedure related to the selection
   of the model;
   
7. handle the problem of autocorrelation and multicollinearity;
   
8. analyse and interpret categorical data using a log-linear
   model.
   

TOPICS:

 Description                                                    Weighting(%)

Multiple Regression: model specification, least                       35.00
squares estimators, maximum likelihood estimators,
Interval estimation, prediction, analysis of variance,
co-efficient of determination, multiple and partial
correlation, tests of hypotheses, test of goodness of fit

Model Selection and Checking:  selection of best                      20.00
model, residual analysis, transformations for variance
stabilisation, diagnostic checking

Multicollinearity:  sources and consequences,                         10.00
identification and solution of multicollinear problem,
Ridge regression

Autocorrelation:  sources and consequences,                           10.00
model for first order autocorrelations, tests for the
presence of autocorrelation, estimated generalised
least squares estimator, prediction

Generalised Linear Model: systematic and random                       15.00
component, likelihood functions, link function,
deviance, applications, generalised additive model

Analysis of Categorical Data: log-linear model for                    10.00
contingency model

TEXT and MATERIALS to be PURCHASED:

RECOMMENDED REFERENCE MATERIALS:

Barker, R.J. & Nelder, J.A. 1993, The GLIM System Release 4, NAG, Oxford.

Cox, D.R. 1977, Analysis of Binary Data, Chapman & Hall, London.

Dobson, A.J. 1990, An Introduction to Generalised Linear Models, Chapman
and Hall, London.

Draper, N. & Smith, H. 1981, Applied Regression Analysis, 2nd edn, Wiley,
New York.

Everitt, B.S. 1977, The Analysis of Contingency Tables, Chapman Hall,
London.

Graybill, F.A. 1961, An Introduction to Linear Statistical Models,
McGraw-Hill, New York.

Montgomery, D.C. & Peck, E.A. 1992, Introduction to Linear Regression
Analysis, 2nd edn, Wiley, New York.

Searle, S.R. 1971, Linear Models, Wiley, New York.

STUDENT WORKLOAD REQUIREMENTS:

	ACTIVITY				HOURS
Directed Study                                	50
Private Study                                 	110

ASSESSMENT DETAILS:

No  *F/S Marks     Due        Description                              Wtg(%)    LBL
1   S              21/08/97  ASSIGNMENT 1                              10.00     Y
2   S              19/09/97  ASSIGNMENT 2                              10.00     Y
3   S              17/10/97  ASSIGNMENT 3                              10.00     Y
4   S              END S2    3 HR CLSD BK THEORY/APPLNS EXAM           70.00     N

OTHER REQUIREMENTS:

1    To   obtain   a   pass  in  the  unit,  students  must   perform
     satisfactorily in all aspects of assessment.
2    The due date for assessments is the date by which a student must
     despatch an assignment to the USQ. The onus is on the student to
     provide  proof  of  the  despatch  date,  if  requested  by  the
     Examiner.
3    Students  MUST  retain a copy of all assignments which  must  be
     produced if and when required by the Examiner.
4    Extensions   for  assignment  submission  may  be   granted   in
     extenuating  circumstances. The decision to grant or  refuse  an
     extension is made by the Examiner. Students should be aware that
     an  application  for  an extension does not  guarantee  that  an
     extension will be granted.
5    Students apply for extension by either applying at the  time  of
     submitting  an  assignment  or  applying  in  writing  prior  to
     submitting  an  assignment.  All relevant  documentation  should
     accompany the application.
6    If assignments are submitted after the due date and no extension
     is  granted,  then  a  penalty up to a maximum  of  20%  of  the
     assignment mark for each working day late may apply.
7    No  further assignments will be accepted for assessment purposes
     after  assignments or model solutions have been released, except
     in extenuating circumstances.
8    The decision  of the Dean shall be final on any dispute that may
     arise in the implementation of these guidelines.