Correlation Regression And: Learning Outcomes
Correlation Regression And: Learning Outcomes
Correlation Regression And: Learning Outcomes
Regression and
Correlation
43.1 Regression 2
43.2 Correlation 17
Learning outcomes
You will learn how to explore relationships between variables and how to measure the
strength of such relationships. You should note from the outset that simply establishing
a relationship is not enough. You may establish, for example, a relationship between the
number of hours a person works in a week and their hat size. Should you conclude that
working hard causes your head to enlarge? Clearly not, any relationship existing here is
not causal!
Regression 43.1
Introduction
Problems in engineering often involve the exploration of the relationship(s) between two or more
variables. The technique of regression analysis is very useful and well-used in this situation. This
Section will look at the basics of regression analysis and should enable you to apply regression
techniques to the study of relationships between variables. Just because a relationship exists between
two variables does not necessarily imply that the relationship is causal. You might find, for example
that there is a relationship between the hours a person spends watching TV and the incidence of
lung cancer. This does not necessarily imply that watching TV causes lung cancer.
Assuming that a causal relationship does exist, we can measure the strength of the relationship
by means of a correlation coefficient discussed in the next Section. As you might expect, tests of
significance exist which allow us to interpret the meaning of a calculated correlation coefficient.
' $
• have knowledge of Descriptive Statistics
( 36)
2 HELM (2008):
Workbook 43: Regression and Correlation
®
1. Regression
As we have already noted, relationship(s) between variables are of interest to engineers who may
wish to determine the degree of association existing between independent and dependent variables.
Knowing this often helps engineers to make predictions and, on this basis, to forecast and plan.
Essentially, regression analysis provides a sound knowledge base from which accurate estimates of
the values of a dependent variable may be made once the values of related independent variables are
known.
It is worth noting that in practice the choice of independent variable(s) may be made by the engineer
on the basis of experience and/or prior knowledge since this may indicate to the engineer which
independent variables are likely to have a substantial influence on the dependent variable. In summary,
we may state that the principle objectives of regression analysis are:
(a) to enable accurate estimates of the values of a dependent variable to be made from known
values of a set of independent variables;
(b) to enable estimates of errors resulting from the use of a regression line as a basis of
prediction.
Note that if a regression line is represented as y = f (x) where x is the independent variable, then
the actual function used (linear, quadratic, higher degree polynomial etc.) may be obtained via the
use of a theoretical analysis or perhaps a scatter diagram (see below) of some real data. Note that
a regression line represented as y = f (x) is called a regression line of y on x .
Scatter diagrams
A useful first step in establishing the degree of association between two variables is the plotting of a
scatter diagram. Examples of pairs of measurements which an engineer might plot are:
If there exists a relationship between measured variables, it can take many forms. Even though an
outline introduction to non-linear regression is given at the end of this Workbook, we shall focus on
the linear relationship only.
In order to produce a good scatter diagram you should follow the steps given below:
1. Give the diagram a clear title and indicate exactly what information is being displayed;
HELM (2008): 3
Section 43.1: Regression
Examples of scatter diagrams are shown below.
+ +
+
+++ +
+ +
+ + +
+ +
+ + + ++ + +
+ +
+ +
+ + +
+ + +
+ ++
Figure 1 Figure 2 Figure 3
Figure 1 shows an association which follows a curve, possibly exponential, quadratic or cubic;
Figure 2 shows a reasonable degree of linear association where the points of the scatter diagram
lie in an area surrounding a straight line;
Figure 3 represents a randomly placed set of points and no linear association is present between
the variables.
Note that in Figure 2, the word ‘reasonable’ is not defined and that while points ‘close’ to the
indicated straight line may be explained by random variation, those ‘far away’ may be due to assignable
variation.
The rest of this Section will deal with linear association only although it is worth noting that techniques
do exist for transforming many non-linear relationships into linear ones. We shall investigate linear
association in two ways, firstly by using educated guess work to obtain a regression line ‘by eye’ and
secondly by using the well-known technique called the method of least squares.
4 HELM (2008):
Workbook 43: Regression and Correlation
®
Pn (xn , yn )
y
y = a + bx
P1 (x1 , y1 )
Qn
Q2
Q1
P2 (x2 , y2 )
O x
Figure 4
The vertical distances of each point from the line y = a + bx are easily calculated as
y1 − a − bx1 , y2 − a − bx2 , y3 − a − bx3 ··· yn − a − bx4
These distances are squared to guarantee that they are positive and calculus is used to minimise the
sum of the squared distances. Effectively we are minimizing the sum of a two-variable expression
and need to use partial differentiation. If you wish to follow this up and look in more detail at
the technique, any good book (engineering or mathematics) containing sections on multi-variable
calculus should suffice. We will not look at the details of the calculations here but simply note that
the process results in two equations in the two unknowns m and c being formed. These equations
are:
X X X
xy − a x−b x2 = 0 (i)
and
X X
y − na − b x=0 (ii)
The second of these equations (ii) immediately gives a useful result. Rearranging the equation we
get
P P
y x
−a−b = 0 or, put more simply ȳ = a + bx̄
n n
where (x̄, ȳ) is the mean of the array of data points (x1 , y1 ), (x2 , y2 ), · · · , (xn , yn ).
This shows that the mean of the array always lies on the regression line. Since the mean is easily
calculated, the result forms a useful check for a plotted regression line. Ensure that any regression
line you draw passes through the mean of the array of data points.
Eliminating a from the equations gives a formula for the gradient b of the regression line, this is:
HELM (2008): 5
Section 43.1: Regression
P P P
xy x y
− Sxy
b = Pn 2 nP n 2 often written as b=
x x Sx2
−
n n
The quantity Sx2 is, of course, the variance of the x-values. The quantity Sxy is known as the
covariance (of x and y) and will appear again later in this Workbook when we measure the degree
of linear association between two variables.
Knowing the value of b enables us to obtain the value of a from the equation ȳ = a + bx̄
Key Point 1
Least Squares Regression - y on x
The least squares regression line of y on x has the equation y = a + bx, where
P P P
xy x y
−
b = Pn 2 nP n 2 and a is given by the equation a = ȳ − bx̄
x x
−
n n
It should be noted that the coefficients b and a obtained here will give us the regression line of y on
x. This line is used to predict y values given x values. If we need to predict the values of x from
given values of y we need the regression line of x on y. The two lines are not the same except in the
(very) special case where all of the points lie exactly on a straight line. It is worth noting however,
that the two lines cross at the point (x̄, ȳ). It can be shown that the regression line of x on y is
given by Key Point 2:
Key Point 2
Least Squares Regression - x on y
The regression line of x on y is
x = a0 + b 0 y
where
P P P
xy x y
−
b0 = Pn 2 nP n and a0 = x̄ − b0 ȳ
2
y y
−
n n
6 HELM (2008):
Workbook 43: Regression and Correlation
®
Example 1
A warehouse manager of a company dealing in large quantities of steel cable needs
to be able to estimate how much cable is left on his partially used drums. A
random sample of twelve partially used drums is taken and each drum is weighed
and the corresponding length of cable measured. The results are given in the table
below:
Weight of drum and cable (x) kg. Measured length of cable (y) m.
30 70
40 90
40 100
50 120
50 130
50 150
60 160
70 190
70 200
80 200
80 220
80 230
Find the least squares regression line in the form y = mx + c and use it to predict
the lengths of cable left on drums whose weights are:
(i) 35 kg (ii) 85 kg (iii) 100 kg
In the latter case state any assumptions which you make in order to find the length
of cable left on the drum.
Solution
P P 2 P P
Excel calculations give x = 700, x = 44200, y = 1860 xy = 118600 so that
the formulae
P P P
xy x y
−
b = Pn 2 nP n 2 and a = ȳ − bx̄
x x
−
n n
give a = −20 and b = 3. Our regression line is y = −20 + 3x, so y = 3x − 20.
Hence, the required predicted values are:
y35 = 3 × 35 − 20 = 85 y85 = 3 × 85 − 20 = 235 y100 = 3 × 100 − 20 = 280
all results being in metres.
To obtain the last result we have assumed that the linearity of the relationship continues beyond
the range of values actually taken.
HELM (2008): 7
Section 43.1: Regression
Task
An article in the Journal of Sound and Vibration 1991 (151) explored a possible
relationship between hypertension (defined as blood pressure rise in mm of mer-
cury) and exposure to noise levels (measured in decibels). Some data given is as
follows:
Noise Level (x) Blood pressure rise (y) Noise Level (x) Blood pressure rise (y)
60 1 85 5
63 0 89 4
65 1 90 6
70 2 90 8
70 5 90 4
70 1 90 5
80 4 94 7
90 6 100 9
80 2 100 7
80 3 100 6
Your solution
8 HELM (2008):
Workbook 43: Regression and Correlation
®
Answer
(a) Entering the data into Microsoft Excel and plotting gives
Blood Pressure increase versus recorded sound level
HELM (2008): 9
Section 43.1: Regression
V(Y |x) = V(α + βx) + V(e) = σ 2 since V(α + βx) = 0.
This implies that for each value of x, Y is distributed with mean α + βx and variance σ 2 . Hence
when the variance is small the observed values of Y will be close to the regression line and when the
variance is large, at least some of the observed values of Y may not be close to the line. Note that
the assumption that the errors e are distributed with mean 0 may be made without loss of generality.
If the errors had any other mean, we could subtract it and then add the mean to the value of c. The
ideas are illustrated in the following diagram.
E (y|x) = α + βx
ei
yi
O x1 x2 x3 xi x
Figure 5
The regression line is shown passing through the means of the distributions for the individual values
of x. The value of y corresponding to the x-value xi can be represented by the equation
yi = α + βxi + ei
where ei is the error of the observed value of y, that is the difference from its expected value, namely
E(Y |xi ) = µy|xi = α + βxi
Now, if we estimate α and β with a and b, the residual, or estimated error, becomes
êi = yi − a − bxi
so that the sum of the squares of the residuals is given by
X X
S= ê2i = (yi − a − bxi )2
and we may minimize the quantity S by using the method of least squares as before. The mathe-
matical details are omitted as before and the equations obtained for b and a are as before, namely
P P P
xy x y
−
b = Pn 2 nP n 2 and a = ȳ − bx̄.
x x
−
n n
Note that since the error ei in the ith observation essentially describes
P 2 the error in the fit of the
model to the ith observation, the sum of the squares of the errors ei will now be used to allow us
to comment on the adequacy of fit of a linear model to a given data set.
10 HELM (2008):
Workbook 43: Regression and Correlation
®
Adequacy of fit
We now know that the variance V(Y |x) = σ 2 is the key to describing the adequacy of fit of our
simple linear model. In general, the smaller the variance, the better the fit although you should note
that it is wise to distinguish between ‘poor fit’ and a large error variance. Poor fit may suggest, for
example, that the relationship is not in fact linear and that a fundamental assumption made has been
violated. A large value of σ 2 does not necessarily mean that a linear model is a poor fit.
It can be shown that the sum of the squares of the errors say SSE can be used to give an unbiased
estimator σ̂ 2 of σ 2 via the formula
SSE
σ̂ 2 =
n−p
where p is the number of independent variables used in the regression equation. In the case of simple
linear regression p = 2 since we are using just x and c and the estimator becomes:
SSE
σ̂ 2 =
n−2
The quantity SSE is usually used explicitly in formulae whose purpose is to determine the adequacy
of a linear model to explain the variability found in data. Two ways in which the adequacy of a
regression model may be judged are given by the so-called Coefficient of Determination and the
Adjusted Coefficient of Determination.
(a) 0 ≤ R2 ≤ 1
(b) a large value of R2 does not necessarily imply that a model is a good fit;
(c) adding a regressor variable (simple regression becomes multiple regression) always in-
creases the value of R2 . This is one reason why a large value of R2 does not necessarily
imply a good model;
(d) models giving large values of R2 can be poor predictors of new values if the fitted model
does not apply at the appropriate x-value.
Finally, it is worth noting that to check the fit of a linear model properly, one should look at plots of
residual values. In some cases, tests of goodness-of-fit are available although this topic is not covered
in this Workbook.
HELM (2008): 11
Section 43.1: Regression
The adjusted coefficient of determination
2
Denoted (often) by Radj , the Adjusted Coefficient of Determination is defined as
2 SSE /(n − p)
Radj =1−
SST /(n − 1)
where p is the number of variables in the regression equation. For the simple linear model, p = 2
since we have two unknown parameters in the regression equation, the intercept c and the coefficient
m of x. It can be shown that:
2
(a) Radj is a better indicator of the adequacy of predictive power than R2 since it takes into
account the number of regressor variables used in the model;
2
(b) Radj does not necessarily increase when a new regressor variable is added.
Both coefficients claim to measure the adequacy of the predictive power of a regression model and
their values indicate the proportion of variability explained by the model. For example a value of
R2 or 2
Radj = 0.9751
may be interpreted as indicating that a model explains 97.51% of the variability it describes. For
example, the drum and cable example considered previously gives the results outlined below with
R2 = 96.2 and 2
Radj = 0.958
2
In general, Radj is (perhaps) more useful than R2 for comparing alternative models. In the context
of a simple linear model, R2 is easier to interpret. In the drum and cable example we would claim
that the linear model explains some 96.2% of the variation it describes.
12 HELM (2008):
Workbook 43: Regression and Correlation
®
Task
Use the drum and cable data given in Example 1 (page 7) and set up a spreadsheet
to verify the values of the Coefficient of Determination and the Adjusted Coefficient
of Determination calculated on page 12.
Your solution
Answer
As per the table on page 12 giving R2 = 0.962 and Radj
2
= 0.958.
Key Point 3
Significance Test for Regression
The test statistic is
SSR
Ftest =
SSE /(n − 2)
where SSR = SST − SSE and rejection at the 5% level of significance occurs if
Ftest > f0.05,1,n−2
Note that we have one degree of freedom since we are testing only one parameter (m) and that n
denotes the number of pairs of (x, y) values. A set of tables giving the 5% values of the F -distribution
is given at the end of this Workbook (Table 1).
HELM (2008): 13
Section 43.1: Regression
Example 2
Test to determine whether a simple linear model is appropriate for the data previ-
ously given in the drum and cable example above.
Solution
We know that
SST = SSR + SSE
( y)2
X P
2
where SST = y − is the total sum of squares (of y) so that (from the spreadsheet
n
above) we have:
SSR = 31500 − 1200 = 30300
Hence
SSR 30300
Ftest = = = 252.5
SSE /(n − 2) 1200/(12 − 2)
From Table 1, the critical value is f0.05,1,10 = 241.9.
Hence, since Ftest > f0.05,1,10 , we reject the null hypothesis and conclude that β 6= 0.
Regression curves
The Section should be regarded as introductory only. The reason for including non-linear regression
is to demonstrate how the method of least squares can be extended to deal with cases where the
relationship between variables is, for example, quadratic or exponential.
A regression curve is defined to be the curve passing through the expected value of Y for a set of
given values of x. The idea is illustrated by the following diagram.
f (y)
Regression curve of y on x
0 x1 x2 x3 xn x
Figure 6
14 HELM (2008):
Workbook 43: Regression and Correlation
®
HELM (2008): 15
Section 43.1: Regression
∂f X
= −2 ebx (y − aebx )
∂a
∂f X
= −2 axebx (y − aebx )
∂b
At a minimum we require
∂f ∂f
= =0
∂a ∂b
which results in the two non-linear equations
X X
yebx − a e2bx = 0
X X
xyebx − a xe2bx
which can be solved by iterative methods to give the values of a and b.
Note that it is possible to combine (for example) linear and exponential regression to obtain a
regression equation of the form
y = (α + βx)eγx
The method of least squares may then be used to find estimates a, b, c of α, β, γ.
16 HELM (2008):
Workbook 43: Regression and Correlation