W/DV Val/rm at/omen mink
Regression Modelling 2013-2014
Lab 2: Condence and Prediction Intervals, model building
Part 1: INTRODUCTION
In the lectures, the use of general formulae has been illustrated for the construction of
a:
(a) 95K: go_n_f_idenc.e int
Lab 2, Regression modelling 2013-2014
Sample solution
Example 1
The pairs command produces a matrix plot as shown below. It shows a strong linear relationship
between log volume and log diameter, and a weaker relationship with log height
Example 3: Fuel Consumption
The data used here were originally collected in a study of fuel consumption in the USA.
Of primary interest is the relationship between the tax rate and fuel consumption,
although various other explanatory variables were also r
PROBABILITY (LEVEL M)
Specimen Class Test
Candidates should attempt all questions. The total number of marks for the paper is 40.
Time allowed 60 minutes.
The continuous random variable U has the U(0, 1) distribution with the following
probability density