Statistics First Year Intitial Practical
Statistics First Year Intitial Practical
Statistics First Year Intitial Practical
14
1 Data
1.1 Submission and compilation of data
Your group will have submitted a small dataset generated during a lab
consisting of mixing different amounts of water at different temperatures, all
measured by the group. The data sets from the groups has been compiled by
the course team into a single excel file available on Vula, under the name
Practical 1.xlsx.
1.2 Tasks
You are to complete two main tasks:
1. Fit a regression model to the data in Practical 1.xlsx, compute the
residuals and check the adequacy of the model using a normal
probability plot of the residuals and a plot of residuals versus predicted
values.
2. Report on the regression model in a written report, including the
adequacy check (the two plots). You should comment on the plots to
the best of your ability, but not ramble.
1.2.1 Fit a regression model, compute and analyse the residuals
An empty excel spreadsheet is provided on Vula, named Analysis of Practical
1.xlsx. This file already contains everything you need to complete the
necessary computations. All you need to do is to paste the data correctly into
this file. Use this file to complete the necessary computations, if you wish. Note
that you will have to copy, paste and sort the residuals appropriately to
obtain the plots.
You may also use any software that is capable of fitting a multiple linear
regression model and provide a normal probability plot of residuals as well as
a plot of residuals versus predicted values. You will also need to compute the
coefficient of multiple determination to satisfactorily analyse the fitted model.
1.2.2 The report
The report for Practical 1 is due on Tuesday 7 October at 08:00. It may be
submitted at the Department of Statistics Reception prior to that time, or to
the lecturer directly only during the 15 minute period preceeding the 08:00
lecture on the day. Each student must submit an individual report.
The report should be no longer than 5 pages including title page and
declaration of non-plagiarism page. The title page must include your name,
student number and degree programme.
2
STA1008S
2014
PRACTICAL
1
Do not include a table of contents. You should use the following heading
structure:
Background
Describe the experiment briefly and describe the data that was generated
(do not copy and paste the whole data set into your report).
Method
Describe the response variable and regressor variables as well as the model
that is going to be fit to the data.
Results
Give the resulting fitted model, the residual plots and the coefficient of
multiple determination.
Analysis
Compare the model to the theoretical results obtained for your group. This
comparison is not prepared in the supplied spreadsheet on Vula. Comment
on the residual plots and coefficient of multiple determination, clearly
expressing any concerns you might have about the model adequacy.
Conclusion
Write your conclusion regarding the results of the regression analysis you just
performed.