TWO VARIABLE ANALYSIS
We often want to know how two variables areassociatedorrelated. We want to know whether an increase
in one variable results in an increase or a decrease in the other.
To analyse the relationship between two variables, we first need to decide which is thedependentvariable
and which is theindependentvariable.
The value of the dependent variabledependson the value of the independent variable.
Next we plot known points on ascatter diagram. The independent variable is placed on the horizontal axis,
and the dependent variable is placed on the vertical axis.
Consider the following two typical scatter diagrams:
In the first scatter diagram the points are quite random.
It is hard to tell how they could be related.
In the second scatter diagram the points are all close
to the red line shown. We say that there is a strong
linear connection orlinear correlationbetween these
two variables. The red line is called theline of best fit
because it best represents the data.
The scatter diagram for theOpening Problemis drawn
alongside. Height is the independent variable and is
represented on the horizontal axis. We see that in general,
as the height increases, the weight increases also.
The weight of a
person is usually
dependenton
their height.
Correlationis a measure of the strength of the relationship or association between two variables.
When we analyse the correlation between two variables, we should follow these steps:
Step 1: Look at the scatter diagram for anypattern.
For a generally upward shape we say that the
correlation ispositive.
As the independent variable increases, the dependent
variable generally increases.
For a generally downward shape we say that the
correlation isnegative.
As the independent variable increases, the dependent
variable generally decreases.
A CORRELATION [11.9]
weight
height
dependent variable
independent variable
75
80
85
90
95
100
105
175 180 185 190 195 200 205
Weight versus Height
height (cm)
weight (kg)
456 Two variable analysis (Chapter 22)
IGCSE01
cyan magenta yellow black
(^05255075950525507595)
100 100
(^05255075950525507595)
100 100
Y:\HAESE\IGCSE01\IG01_22\456IGCSE01_22.CDR Monday, 27 October 2008 2:14:32 PM PETER