Vous êtes sur la page 1sur 26

Correlation

 In general, a positive linear relationship (or


correlation) between two variables, X and Y, occur
when high values on X tend to be associated with high
values on Y, and conversely, low values on X tend to be
associated with low values on Y.
 Also, a negative linear relationship (or correlation)
between two variables, X and Y, occur when high
values on X tend to be associated with low values on Y,
and conversely, low values on X tend to be associated
with high values on Y.

2
Correlation
 In previous researches, positive correlations have been
found , for example, between
 Motivation and Academic success

 Crime rate and Unemployment rate

 Salary and job satisfaction

 Verbal ability and proficiency test performance

 Teacher quality and student success

3
Correlation
 In previous researches, negative correlations have been
found , for example, between
 Anxiety and test performance

 College stress and academic success

 Social support and isolation

 Education of Parents and Number of children

4
Correlation Analysis
 Analysis of the relationships among variables is called
Correlation Analysis. It includes the measurement of the
correlation between the variables. It is only concerned
with the strength of the relationship; no causal effect is
implied.

 A scatter plot (or scatter diagram) can be used to show


the relationship between two numerical variables

 Correlation coefficient is used to measure the direction


and strength of the linear relationship between variables.

5
The value of r
 It ranges from -1 to +1

-1 0 +1

 Positive and negative signs indicate the direction of


the relationship.

6
When r is 0:
 A correlation coefficient of 0 indicates that there is
no linear relationship between two variables X and Y.
 However, there might be some nonlinear relationship
between them.
 Nonlinear relationships can either be quadratic,
logarithmic, exponential, etc.
 A scatter diagram can give us a picture of nonlinear
relationships.

7
The Correlation Coefficient
 Unit free
 Ranges between –1 and 1
 The closer to –1, the stronger the negative
linear relationship
 The closer to 1, the stronger the positive
linear relationship
 The closer to 0, the weaker any linear
relationship
The Correlation Coefficient
Y Y Y

X X X
r = -1 r = -.6 r=0

Y Y

X X
X
Statistics for Managers Using
Microsoft Excel, 5e © 2008 r = +1 r = +.3
Pearson Prentice-Hall, Inc.
Absolute value of r
The absolute value of r indicates the
strength of the linear relationship.
 The closer to 1.0 the stronger the linear
relationship
 The closer to zero the weaker the linear
relationship.

10
Sample Data with r =1.0
Data set A

X Y
1 2
2 4
3 6
4 8
5 10
11
Sample Data with r close to 1
Data set B

X Y
1 2.5
2 3.5
3 5
4 7
5 9
12
Sample Data with r close to 0
Data set E

Scatter Plot of Data set E


X Y with its Estimated Line
(r = 0.04)

1 1 10
8
2 10 Y
6
4

3 7 2
0
0 1 2 3 4 5 6
4 9 X

5 2

13
Guide Interpretation
(Note: Rule of Thumb only)
Absolute value of r Interpretation

0.00 – 0.29 Very Low

0.30 – 0.49 Low

0.50 – 0.69 Moderate

0.70 – 0.89 High

0.90 – 1.00 Very High

14
Pearson Product Moment Correlation Coefficient
(or simply “Pearson r” or “r”)

Cov( X, Y )
r
Var ( X ) Var ( Y )
where: X and Y are variables under consideration.

15
Testing Pearson r for
significance
 While the Pearson coefficient provides the
information about the strength and direction of the
linear relationship, it is very important to determine
whether there is a linear relationship between the
variables in the entire population to which the sample
belongs.
 If the correlation coefficient is not statistically
significant, there is no evidence of a linear relationship
between X and Y in the population to which the
sample belongs.

16
Test for significance of r
 Null and alternative hypotheses
 H0: β1 = 0 (no linear relationship)
 H1: β1 ≠ 0 (linear relationship does exist)
 Test statistic

d.f.  n  2

Statistics for Managers Using Microsoft Excel, 5e © 2008


Prentice-Hall, Inc.
Four Requirements
for Pearson Correlation
#1: The two variables should be measured
at the interval or ratio level (i.e., they are
continuous).
#2: There needs to be a linear relationship
between the two variables.
If the relationship displayed in your scatterplot is
not linear, you will have to either use a non-
parametric equivalent to Pearson’s correlation or
transform your data.

18
The following scatterplots highlight the potential
impact of outliers:

Outliers are simply few data points within your data that
do not follow the usual pattern.

19
Use the following to test for the normality of
the variables:

• Kolmogorov-Smirnov Test
• Wilk-Shapiro Test

20
Coefficient of Determination
 Coefficient of determination = r2
 r2 = proportion of variance shared by two variables.
 It indicates what proportion of the variance in one of
the correlated variables is associated with the variance
in the other variable.
 If r = .30 between motivation and task performance,
then r is 0.09. This means that 9% of an individual’s
differences in motivation is associated with the
differences in task performance (or vice versa).

21
Steps in Correlation Analysis
Step 1: Formulate the Hypothesis
Ho: There is no significant relationship between X and
Y.
Ha: There is a significant relationship between X and Y.

Step 2: Specify the Level of significance

22
Steps in Correlation Analysis
Step 3: Gather Data
Step 4: Check the necessary assumptions.
Step 5: Compute appropriate correlation
coefficient and test for significance.
Step 6: Make a decision.

23
Example
Research Question: Is there a significant relationship
between a firm’s financial performance and the age of
the chief financial officer?

Null Hypothesis: There is no significant relationship


between a firm’s financial performance and the age of
the chief financial officer.

24
Using Microsoft Excel

Debt-to-Asset
Age Ratio
Age 1
Debt-to-Asset
Ratio 0.3092 1
25
Using SPSS

26

Vous aimerez peut-être aussi