CHAPTER 10

PEARSON'S CORRELATION ANALYSIS

10.1 Introduction

This chapter discusses Pearson's correlation analysis covering the concept, how to analyze and
interpret the output. The Pearson correlation coefficient is a measure of the strength of the linear
equation between two variables. Basically, Pearson's correlation tries to draw a line of best fit to
the data of two variables to show the extent to which all the data points can form a line of best fit.

Correlation analysis is used to measure the strength of the relationship between two continuous
variables, namely variables at the measurement level of intervals and ratios. For example, scores
in tests, body weight and so on. Pearson's correlation coefficient takes values that are +1 and -1,
where +1 is a perfect positive correlation, and -1 is a perfect negative correlation. 0 means there is
no linear correlation between the two variables. The symbol rxy shows the correlation between
variables x and y. If the data is ordinal level, then you can analyze the data using Spearman's
correlation or Kendall's Tau correlation but cannot use Pearson's correlation.

Interpreting Correlation Coefficients

The correlation coefficient is an index that shows the degree of strength of the relationship
between two variables. We can summarize the conclusions about these values as below:

Correlation Coefficient Value Relationship

1.00 The perfect relationship
0.80 – 0.99 Very strong
0.60 – 0.79 Strong
0.40 – 0.59 Medium
0.20 – 0.39 Weak
0.01 - 0.19 Very weak
0.00 No relationship

Testing hypotheses about correlations,

The null and alternative hypotheses in this test are:

55
 H 0 : rxy = 0 (there is no linear relationship between variables X and Y)
H 1 : rxy  0 (there is a linear relationship between variables X and Y)

The assumptions for the Pearson correlation analysis are as follows;

1. Both variables are continuous (interval level or ratio)

2. Linear relationship between the two variables
3. A random sample of data taken from the population
4. The values for the data of the two variables are not related or independent
5. Variables are normally distributed or close to normal
6. No extreme data (outliers)

10.2 SPSS Procedure for Correlation Analysis

Instructions: Open Data Chapter 10_Correlation 1

A researcher wants to study whether self-esteem has a strong and significant positive relationship
with a person's happiness. A total of 50 employees in an organization were selected as a sample to
answer the distributed questionnaires. The completed questionnaires will be analyzed using the
following steps;

1. Klik Analyze > Correlate > Bivariate

56
2. Click the variables Self-esteem and Happiness . Enter in the Variables box.

Make sure Pearson is selected from the Correlation Coefficients box and Two-tailed is
selected from the Test of Significance box.

Check the Flag significant correlations box.

See display below:

2. Click Options
3. Click Means and standard deviation in the Statistics box

See display below

57
 2. Klik Continue
3. Klik OK

Descriptive Statistics
Mean Std. Deviation N
Self-esteem 4.14 2.030 50
Happiness 3.98 1.684 50

The descriptive statistics table above shows the number of respondents for this study as many as
50 people. The mean value for the self-esteem variable is 4.14 with a standard deviation of 2.03.
The mean value for the happiness variable is 3.98 with a standard deviation of 1.684. These surveys
use a 7-point likert scale.

58
 Correlations
Self-esteem Happiness
Self-esteem Pearson Correlation 1 .801**
Sig. (2-tailed) <.001
N 50 50
Happiness Pearson Correlation .801** 1
Sig. (2-tailed) <.001
N 50 50
**. Correlation is significant at the 0.01 level (2-tailed).

Description

The table above shows the value of the correlation coefficient is +0.801. How do we interpret the
correlation value of +0.801? First, the positive sign shows a direct relationship between the self-
esteem variable and the happiness variable. The higher a person's self-esteem, the higher their
happiness. Asterisk symbols (**) indicate significance at the 0.01 significance level (ie 99%
confidence interval). The value of 0.801 is very close to the value of 1.00 (perfect positive
correlation) showing that the correlation between self-esteem and happiness is a very strong and
significant positive relationship.

EXAMPLE 2:

Instructions: Open Data Chapter 10_Correlation 2

This Pearson correlation analysis is to determine the strength of the relationship between the
variable depression and job satisfaction for a group of employees in an organization.

Instructions:

Please follow steps 1 to 6 as above.

1. Click Analyze > Correlate > Bivariate

2. Click the depression and job satisfaction variables. Enter in the Variables box.

59
Make sure Pearson is selected from the Correlation Coefficients box and Two-tailed is selected
from the Test of Significance box.

Check the Flag significant correlations box

3. Click Options
4. Mark the Means and standard deviation in the Statistics box
5. Click Continue
6. Click OK

See the SPSS output as below;

Descriptive Statistics
Mean Std. Deviation N
Depression 4.14 2.043 42
Job satisfaction 4.29 1.715 42

Correlations
Depression Job satisfaction
Depression Pearson Correlation 1 -.826**
Sig. (2-tailed) <.001
N 42 42
Job satisfaction Pearson Correlation -.826** 1
Sig. (2-tailed) <.001
N 42 42
**. Correlation is significant at the 0.01 level (2-tailed).

60
Description

The table above shows the value of the correlation coefficient is -0.826. The negative sign indicates
an inverse relationship between the depression variable and the job satisfaction variable. The
higher a person's level of depression, the less their job satisfaction. The value of -0.826 approaches
the value of -1.00 (perfect negative correlation). We can conclude that the correlation between
depression and job satisfaction is a very strong and significant negative relationship.

SUMMARY

Pearson's correlation analysis is used to measure the strength of the relationship between two
continuous variables, namely variables at the measurement level of interval and ratio. Pearson's
correlation coefficient is either in the form of positive correlation, negative correlation or no
correlation between the studied variables.

61