1

MEASURES OF RELATIONSHIP

CORRELATION
 It is a relationships or association existing between two variables.
 Correlation analysis is a method used to measure the strength of relationship between two or more variables.
 Relationship could be roughly estimated using the scatter diagram. This could be done by plotting the point
corresponding to each observation on a Cartesian plane. If points closely approximate a line , there is a
relationship.

TYPES OF CORRELATION
 Positive Correlation
 this exists when high scores in one variable are associated with high scores in the second variable and vice
versa.
 Points in the scatter diagram closely follow a straight line rising to the right
 Negative Correlation
 This exists when high scores in one variable are associated with low scores in the other and vice versa.
 Points in the scatter diagram closely follow a straight line falling to the right
 Zero Correlation
 Scores in one variable tend to score neither systematically high nor low in the other variable
 The points in the scatter diagram are spread in a random manner.

CORRELATION COEFFICIENTS
 This is a descriptive measure that describes the magnitude or degree of relationship between two variables
 Usually denoted by r, value of which ranges from -1 to +1

Value of r Interpretation
± 1.00 Perfect correlation
± 0.80 to ± 0.99 High correlation
± 0.60 to ± 0.79 Moderately high correlation
± 0.40 to ± 0.59 Moderate correlation
± 0.20 to ± 0.39 Low correlation
± 0.01 to ± 0.19 Negligible correlation

PEARSON PRODUCT-MOMENT CORRELATION COEFFICIENT

 Also called Pearson-r, in honor of British statistician Carl Pearson who derived the formula
 appropriate when data are interval or ratio scale
 Steps
1. Square x and call it x2.
2. Square y, call it y2.
3. Multiply x and y. Call it xy.
4. Find x, y, xy, x2,  y2.
5. Use the formula:

Nxy – (x) ( y)

r = --------------------------------------------------
[ Nx2 - ( x)2] [Ny2 - ( y)2]

6. Interpret obtained results.

 Test of Significance
 when two variables are related, it is imperative to determine whether the relationship is significant or not. If
it is significant, there really is a relationship. Otherwise, the relationship is simply by chance.
 Steps in testing significance.
1. Determine degrees of freedom using the formula df = n – 2 , where n is the number of paired
observations.
2. Get tabular value using Critical Values of the Correlation Coefficient.
3. Compare tabular value and the computed value.
If valuetab < valuecomp , then r is significant
 2

 Coefficient of Determination
 this tells the amount of variation in Y that could be accounted for by the variation in X.
 this could be obtained by squaring the value of r and is denoted by r2.
Ex. – r = 0.8, r2 = 0.64, means that 64% of the variation in Y is attributed or due to variations in x

 Illustrative Example.
A businessman would like to determine if there is a relationship between the size of a store and the profit to
be earned. The following data gathered from 8 storeowners pertain the size of the store in sq. m and the profit earned
in thousand pesos.
a. Calculate r and interpret.
b. At 0.01 level, is there a significant relationship between the store size and the profit earned?

Store X Y X2 Y2 XY
Store Size Profit
A 35 20
B 22 15
C 27 17
D 16 9
E 28 16
F 12 7
G 40 22
H 32 23


SPEARMAN RANK- ORDER CORRELATION COEFFICIENT

 appropriate when data are ordinal scale (ranks)
 formula was derived by British psychologist Charles Spearman
 symbol used is  (rho)
 steps
i. Rank scores in column x and y. Call them Rx and Ry respectively.
ii. Get the difference between ranks. Call it D
iii. Square each difference. Call it D2. Find D2
iv. Use the formula:

6 D2
 = 1 - -------------
N (N2 – 1)

v. Interpret obtained result

 Testing Significance
 same as Pearson r except that the table to be used should be the critical Values of the
Spearman Rank Correlation Coefficient

 Illustrative Example

Store X Y Rx Ry D D2
Store Size Profit
A 35 20 2 3 -1 1
B 22 15 6 6 0 0
C 27 17 5 4 1 1
D 16 9 7 8 -1 1
E 28 16 4 5 -1 1
F 12 7 8 7 1 1
G 40 22 1 2 -1 1
H 32 23 3 1 2 4
 0 10
 3

PHI - COEFFICIENT
 appropriate for nominal dichotomous variables
 formula was also derived by Carl Pearson in 1901
 Phi coefficient table

1 2
1 a b a+b
2 c d c+d
a+c b+d

 Formula:
ad – bc
r = ----------------------------------------
 (a + b) (c + d) (a + c) (b + d)

 Illustrative Example:
Let x represent sex , (1) male and (0) female, and let y represent opinion toward Charter Change, for (1)
and against (0). Ten male respondents were for the ChaCha while 6 were against. Likewise, 5 females were for the
Chacha while 12 were against. Find the degree of correlation between sex and opinion in the chacha.

Sex ( x )
0 1
0 12 6 18
1 5 10 15
17 16

(12) (10) – (5) (6)

r = ----------------------------------------
 (18) (15) (17) (16)

POINT BISERIAL CORRELATION

 a special case of the Pearson r
 it correlates a real dichotomous with an interval variable
 Formula:
X 1 - X0 n1 n0
Rpb = ----------- ---------
sx n(n-1)

where : X1 - mean of those labeled 1 in the real dichotomous y

X0 - mean of those labeled 0 in the real dichotomous y
n1 - number of samples labeled 1 in y
n0 - number of samples labeled 0 in y
n - total number of samples n1 + n0
sx - standard deviation of all the x values.

 Illustrative Example:
Five boys (1) and 4 girls (0) took a 20-item test in psychomotor dexterity x. The results
are tabulated below: Compute for the correlation between x and y
Sex Dexterity
Students y x
1 1 12
2 0 10
3 1 20
4 0 15
5 1 15
6 1 12
7 0 18
8 0 10
 4

9 1 7

REGRESSION

 Discusses the formal procedure for predicting one variable from a knowledge of another variable.
 Line of best fit – the line that best represents the set of points in a given data.
 Prediction equation – the equation of the line that best represents the set of points

Ŷ = A + Bx

Nxy – (x) ( y)

B = -------------------------
Nx2 - ( x)2

A = y - Bx

Alternative formula: Sy
Ŷ = rxy ------ ( x – x) + y
Sx

Where : Sy - standard deviation of the y values

Sx - standard deviation of the x values
rxy - correlation coefficient
x - mean of the x values
y - mean of the y values

 Standard error - is the standard deviation of all the errors in the set of data

Se = Sy 1 – rxy 2
Illustrative Example:
Given that the following are scores of 9 students in Science and Math tests, find the
corresponding math scores for a science score of:
a. x = 1
b. x = 3
c. x = 8

Student Science Math X2 Xy

X Y
1 2 3
2 4 4
3 5 4
4 5 5
5 6 6
6 6 7
7 7 8
8 9 8
9 10 9