Scribd
Upload
14 views26 pages

10 CORRELATION Phi-Coefficient

The document discusses the Phi-coefficient, a nonparametric statistic used to measure the strength of association between two dichotomous variables. It outlines the requirements for its use, interpretation of values, and significance testing methods. An example is provided to illustrate its application in analyzing the relationship between marital status and sex at birth.

Uploaded by

baribadk
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
14 views26 pages

10 CORRELATION Phi-Coefficient

The document discusses the Phi-coefficient, a nonparametric statistic used to measure the strength of association between two dichotomous variables. It outlines the requirements for its use, interpretation of values, and significance testing methods. An example is provided to illustrate its application in analyzing the relationship between marital status and sex at birth.

Uploaded by

baribadk
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/ 26

STATISTICAL TOOLS

AND PROCEDURES
FOR TESTING
RELATIONSHIPS
PREPARED BY JE-ANN H. IROG
NON-LINEAR
CORRELATION &
NONPARAMETRIC
PHI-COEFFICIENT
Phi-Coefficient
Also referred to as the mean square contingency coefficient
and denoted by the symbol φ.

Introduced by Karl Pearson, this measure is similar to the


Pearson correlation coefficient in its interpretation.
Phi-Coefficient
The phi-coefficient is one of a number of correlation
statistics developed to measure the strength of
association/relationship between two binary
variables. It is a nonparametric statistic used in cross-
tabulated table data where both (two) variables are
dichotomous. Dichotomous means that there are
only two possible values for a variable.
Phi-Coefficient
We should note that although the phi-coefficient can
be used to assess the relationship between two
dichotomous variables, the more common statistical
procedure is a chi-square statistic, which can be used
for two or more dichotomous variables.
Requirements in Using Phi-Coefficient
1. At least one variable is nominal
2. The distribution is skewed or not normal
3. Two variables that are both dichotomous
4. Two independent variables OR one independent variable
and the other is dependent variable

A dichotomous variable is a variable with only two mutually exclusive response options.
Phi-Coefficient
It intersects variables across a 2x2 matrix to estimate whether there
is a non-random pattern across the four cells in the 2x2 matrix

What does this


mean?
Example Data Set

Subjects Sex at Birth Marital Status


1= Male 1 = Single
2= Female 2 = Married

Two
Dichotomous
Variables
Example Data Set
It intersects variables across a
Subjects Sex at Birth Marital Status 2x2 matrix to estimate
1 = Male 1= Single whether there is a non-
2 = Female 2 = Married random pattern across the
A 1 2 four cells in the 2x2 matrix
B 1 1
C 1 1 SEX AT BIRTH
D 2 2 Male Female
E 2 2 MARITAL Married
F 1 1 STATUS Single
G 2 2
H 2 1
I 2 2
J 1 1
K 1 1
L 2 1
Four cells in the 2x2 matrix
MALE FEMALE

MARRIED 1 4

SINGLE 5 2
Formula

MALE FEMALE

MARRIED 1 (a) 4 (b)

SINGLE 5 (c) 2 (d)


Formula

MALE FEMALE

MARRIED 1 (a) 4 (b) 5 (e)

SINGLE 5 (c) 2 (d) 7 (f)

6 (g) 6 (h)
Interpretation
Similar to a Pearson Correlation Coefficient, a Phi
Coefficient takes on values between -1 and 1 where:
-1 indicates a perfectly negative relationship
between the two variables.
0 indicates no association between the two
variables.
+1 indicates a perfectly positive relationship between
the two variables.
Interpretation
A Phi-coefficient of 0 would indicate that there is no systematic pattern across the 2x2 matrix.

SEX AT BIRTH
Male Female
MARITAL Married 3 3
STATUS Single 3 3

or SEX AT BIRTH
Male Female
MARITAL Married 5 5
STATUS Single 1 1
Interpretation
A positive Phi-coefficient would indicate that most of the data are in the diagonal cells.

SEX AT BIRTH
Male Female
MARITAL Married 4 1
STATUS Single 2 5
Interpretation
A negative Phi-coefficient would indicate that most of the data are in the off-diagonal cells. A

SEX AT BIRTH
Interpretation
Interpretation
In general, the further away a Phi Coefficient is
from zero, the stronger the relationship between
the two variables.
In other words, the further away a Phi Coefficient
is from zero, the more evidence there is for some
type of systematic pattern between the two
variables.
Notes in Interpretation
The signs of negative (-) and positive (+) in
the computed Phi coefficient are not much
important, especially if both of your
variables are nominal. It is better to look at
the pattern in the cross tab to better
interpret and discuss the result.
SEX AT BIRTH Phi-
Coefficient

+.507

There is a strong relationship between marital status and sex at


birth with being male making it more likely you are married and
being female making it more likely you are single.
Testing the Significant of Phi coefficient
To test if the computed Phi coefficient
is significant or not significant, use the
formula below:

N (r)2

Where
N is the sample size
r is the computed Phi coefficient

Then, look at the (Chi-square) X2


distribution table in df 1 (constant).
Testing the Significant of Phi coefficient
df 1
Chi-square table from statology.org
Alpha level 0.05
One-tailed test
If cv > tv; reject null hypothesis
If cv < tv; accept null hypothesis

Testing the significance using one-


tailed test and alpha level 0.05, the
computed value 3.042508 with df 1
has a table/critical value of 3.841.
Since the computed value (3.042) is
less than the critical value (3.841),
accept null hypothesis. Thus, there is
no significant relationship between
the respondents’ belief in UFOs and
Horoscopes.
FOR THE SIGNIFICANCE…
SEX AT BIRTH
Phi-
Coefficient

+.507

Though the computed Phi coefficient value +0.507 indicates that


there is a strong relationship between marital status and sex at
birth with being male making it more likely you are married and
being female making it more likely you are single, in terms of
significance test, the relationship between the two variables is not
significant. This is indicated by a computed value of 3.084588 with
a table value of 3.841 using alpha level 0.05 and one-tailed test.
Since the computed value is less than the table value, the null
hypothesis is accepted.
REFERENCES
Gravetter, F.J. & Wallnau, L.B. (2017, 2013). Chapter 15: Correlation, Statistics for the
behavioral sciences, tenth edition, pp 485-527. USA: Cengage Learning.
Statistical tables. https://onlinelibrary.wiley.com/doi/pdf/10.1002/9781118342978.app2
Phi correlation coefficient. https://methods.sagepub.com/reference/the-sage-encyclopedia-
of-educational-research-measurement-and-
evaluation/i15899.xml#:~:text=The%20phi%20correlation%20coefficient%20(phi,where
%20both%20variables%20are%20dichotomous.
https://www.statisticshowto.com/phi-coefficient-mean-square-contingency-coefficient/
Phi coefficient https://www.statology.org/phi-coefficient/
Phi coefficient https://flowjo.typepad.com/the_daily_dongle/files/Phi-coefficient.pdf
PRACTICE PROBLEM
FAMILY MONTHLY NET INCOME
HIGH (Above Php LOW (Php 5000 to
25000) 25000)
MARITAL STATUS
OFW 26 4
NOT OFW 9 21

INSTRUCTION:

1. State your null hypothesis.


2. Compute for Phi coefficient
3. Test significance at alpha level 0.05 and one-tailed test.
4. Interpret the result.

You might also like