0% found this document useful (0 votes)
25 views40 pages

L7 Correlation

The document discusses different types of correlation coefficients and how to calculate them. It covers Pearson's correlation coefficient, Spearman's rank correlation coefficient, point biserial correlation coefficient, and Phi coefficient. Formulas and steps to compute the coefficients are provided along with examples using Excel and SPSS.

Uploaded by

derisivebrat
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
25 views40 pages

L7 Correlation

The document discusses different types of correlation coefficients and how to calculate them. It covers Pearson's correlation coefficient, Spearman's rank correlation coefficient, point biserial correlation coefficient, and Phi coefficient. Formulas and steps to compute the coefficients are provided along with examples using Excel and SPSS.

Uploaded by

derisivebrat
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 40

Lesson 7:

Correlation
Relationship between Data
• To measure and describe a relationship between two variables we usually use
correlation statistical techniques.
• In this technique, there is no intention of manipulating or controlling the variables
since they exist naturally in environment.
• Using correlation, we may look into the relationship of subject grade to economic
status or distance of the student from the teacher and their subject grade.
• In this sample there is no manipulation of variables, instead this looks simply in the
observation of occurrence.
• There are also only two variables that are observed, normally represented by X
and Y variable, the value of which can be represented in a scatter plot.
• This would allow us to see the pattern or trend of the relation between the x
variables (horizontal axis) and y variable (vertical axis).

Correlation 2
• A variety of Correlation Coefficient has been developed over the years for
different combinations of types of scales. These are the following.

Correlation 3
The Characteristics of a Relationship
• A correlation describes and measures the characteristics of two
variables. There are some characteristics of the relationship.
1. The direction of the relationship. The sign of the correlation that is
either positive or negative would describe the relationship
direction.

a. Positive Relationship. This results when one value increases the


other value also increases – meaning that the two variables are
moving in the same direction. This is the direct correlation or
the positive correlation
Correlation 4
• In this example, as the daily allowance of the students increases and so is
their Grade Point Average. This shows a positive relationship between the
daily allowance and grade point average.
Correlation 5
b. A negative correlation tends to go in the opposite direction. As the x
variable increase the y variable decreases. This is the inverse relationship or
the negative relationship.

• In this example, as the daily allowance of the student increases, their Grade Point
Average decrease. This shows a negative relationship or inverse relationship between the
daily allowance and grade point average.

Correlation 6
• In statistics, the correlation coefficient r measures the strength and direction of a
linear relationship between two variables on a scatterplot. The value of r is always
between +1 and –1. To interpret its value, see which of the following values your
correlation r is closest to:

Correlation 7
Pearson Product Moment of Coefficient of Correlation r
• The Pearson Product Moment of Coefficient of Correlation r is an index of
relationships between two variables. The value of the Pearson could range
from +1 to -1, which is a perfect correlation. If the value is zero (0) then
there is no correlation but if the value you got is more than 1.0 then there
must be something wrong with your computation.

• The formula for the Pearson r is

Correlation 8
Steps in computing for the value of Pearson r
1. Count the number of observation or the n

2. Compute the sum of x that is ∑x

3. Square the value x (x²) and get the sum of x² (∑x²)

4. Compute the sum of y that is ∑y

5. Square the value y (y²) and get the sum of y² (∑y²)

6. Get the product of the x and y and the sum of xy (∑xy)

7. Apply the given formula.

Correlation 9
Example 1:
1. Count the number of
observation or the n

2. Compute the sum of x that


is ∑x

3. Square the value x (x²) and


get the sum of x² (∑x²)
437,425

Correlation 10
Example 1:
4. Compute the sum of y that is ∑y
5. Square the value y (y²) and get the sum of y² (∑y²)

Correlation 11
Example 1:
6. Get the product of the x and y and the sum of xy (∑xy)

Correlation 12
Example 1:
7. Apply the given formula.

• Strong Positive Correlation


Correlation 13
Steps in Computing Pearsons r Using Excel

Correlation 14
Steps in Computing Pearsons r Using Excel

Correlation 15
Steps in Computing Pearsons r Using SPSS

Correlation 16
Steps in Computing Pearsons r Using SPSS

Correlation 17
Spearman Rank Order of Coefficient
Correlation
• The Spearman Rho is a non-parametric test of correlation for it does not
require normality in distribution. It is used to look into the significant
relationship of two variables when ranking is used.
• The formula for Spearman Rho is:

Correlation 18
Steps in Computing Spearman Rho
1. Rank the data in x from highest to lowest

2. Rank the data in y from highest to lowest

3. Compute the Difference between the Rank of x and y

4. Square the difference of the ranks

5. Find the summation of the squared difference

6. Determine the n

7. Use the formula given

Correlation 19
Example 2:
1. Rank the data in x
from highest to lowest

• Create a temporary rank column


and arrange the score from highest
to lowest. The highest score gets
the rank 1. In case of similar scores,
add the RANKS (not the scores)
and divide it by the number of
similar scores. The temporary ranks
will guide you as to what rank is
next.

Correlation 20
Example 2:
2. Rank the
data in y
from
highest to
lowest

Correlation 21
Example 2:
3. Compute the
Difference
between the
Rank of x and y

4. Square the
difference of the
ranks

5. Find the
summation of
the squared
difference

Correlation 22
6 σ 𝐷2
Example 2: 𝑟𝑠 = 1 −
𝑛 𝑛2 − 1

6. Determine the n 6 399.21


𝑟𝑠 = 1 −
15 152 − 1

7. Use the formula given 2395.26


𝑟𝑠 = 1 −
15 225 − 1

2395.26
𝑟𝑠 = 1 −
15 224

2395.26
𝑟𝑠 = 1 −
3360
𝑟𝑠 = 1 − 0.71
𝒓𝒔 = 𝟎. 𝟐𝟗

Correlation 23
Point Biserial Correlation Coefficient (rpbi)
• The point biserial correlation coefficient (rpbi) is a statistics used to estimate the
degree of relationship between a dichotomous and an interval or ratio scale. If a
researcher want to find out the degree of relationships between gender (male and
female) and IQ score (scale) the point biserial correlation coefficient is appropriate.
The interpretation of which is similar to Pearson r – that can range -1 to +1.

• To find the point biserial correlation, the dichotomous or binomial variable is


converted to numerical values. We can assign a value of zero (0) to one category
and a value of one (1) to the other category. Then use the regular Pearson
Correlation formula to convert the data. It is customary to use zero (0) and one
(1) but using other two different numbers would work equally well and would not
affect the value of the correlation.

Correlation 24
Example 3:
• A researcher wants to determine if gender is related to IQ. Since IQ is continuous data
and gender is dichotomous, the point biserial correlation coefficient is appropriate.

• Since the computed value is 0.67, it shows that IQ and Gender has strong relationship.
Correlation 25
Computing Point Biserial Correlation Using Excel
• Enter the data, click the function (fx) cell then select pearson

Correlation 26
Computing Point Biserial Correlation Using Excel

• Select the X data for array 1 and Y data for array 2, then click OK.

Correlation 27
Computing Point Biserial Correlation Using SPSS
• Enter in data, then click analyze, select correlate and bivariate

Correlation 28
Computing Point Biserial Correlation Using SPSS
• Move the data from left box to the right, then click OK

• You can also use this web site to compute http://vassarstats.net/pbcorr.html

Correlation 29
Phi-Coefficient
• This statistics is used for variables that are both nominal and dichotomous. In
this statistics, contingency table is used to present the data. It is a two-
dimensional table containing frequencies by categories. It is more appropriate
for two by two since each variable only take two values.
• As an example, let us consider this data that is organized by gender and
position.

Correlation 30
Phi-Coefficient
• Contingency tables are often coded as
below to simplify calculation of the Phi
coefficient.

Correlation 31
Interpretation of the Phi-Coefficient
• -1.0 to -0.7 strong negative association.

• -0.7 to -0.3 weak negative association.

• -0.3 to +0.3 little or no association.

• +0.3 to +0.7 weak positive association.

• +0.7 to +1.0 strong positive association.

Correlation 32
Similar Linear Regression Analysis
• The simple linear regression analysis is used when there is a relationship
between x and y variables. This is used to predict the value of y given the
value of x. We use linear regression analysis for forecasting and prediction.
• The formula is:

Correlation 33
Similar Linear Regression Analysis
• To solve for the b use the formula below

• To solve for a use the formula below:

𝑎 = 𝑦ത − 𝑏𝑥ҧ

Correlation 34
To solve, here are the steps:
1. Use the data as computed using the Pearson Product Moment
Coefficient of Correlation r or Pearson’s r.
2. Compute for b, by computing the following first: ∑xy, ∑x, ∑y,
∑x²

3. Then solve the b using the formula for b.

4. Compute for a by finding the mean of y, 𝑦ത and then mean of x, 𝑥.ҧ

Correlation 35
Example 4:
• A research looks into the relationship of the number of hours
that the students reviewed (x) and the grades (y) they got in their Statistics
Exam. Find if there is a significant relationship between the number of hours
reviewed and their grades.

Correlation 36
Example 4:

Correlation 37
Example 4:
1. Use the data as
computed using the
Pearson Product Moment
Coefficient of Correlation r
or Pearson’s r.

Correlation 38
Example 4:
2. Compute for b, by computing the following
first: ∑xy, ∑x, ∑y, ∑x²

3. Then solve the b using the formula for b.

Correlation 39
Example 4:
4. Compute for a by finding the mean of y, ´y and then mean of x, ´x.

• Supposed we want to predict the grade (y) of a student who reviewed for 6 hours (y),
get the value of y by computing the simple linear regression. Solve for the a then the y.
a is: y is:

Correlation 40

You might also like