MODULE No.

MEASURES OF
RELATIONSHIPS
 Assessment in Learning 1

MODULE No. 7
MEASURES OF RELATIONSHIPS

I. Objectives:

1. Determine relationships of two sets of scores

2. Calculate two common correlation coefficients:
a) Pearson Product-Moment (Pearson r)
b) Spearman Rank ( r ranks or Spearman rho)

II. Anticipatory Set

Teachers conducting action research sometimes explore correlation of variables of interest,

for example, the relationship between achievement scores in mathematics and science; the relationship
between scores in vocabulary and reading comprehension tests, etc. In these cases, measures of
relationships are called for. In this module, two commonly used measures of relationships are
presented with illustrative examples.

III. Lesson Proper

Measures of relationship are used to indicate the degree to which two sets of scores covary of
are related. We intuitively seek relationships by such statements as: “If high scores in variable X tend to
be associated with high scores on variable Y, then the variables are related,” or “ If high scores on
variable X tend to be associated with low scores on variable Y, then the variables are related.”
Relationships can be either positive or negative and either strong or weak.
We use correlation coefficient as a statistical summary of the nature of the relationship
between two variables. Correlation coefficients provide us with an estimate of the quantitative degree
of relationship. The correlation coefficient ranges from -1.0 to + 1.0. Values close to -1.0 or +1.0
indicate a strong linear relationship. The closer the correlation is to zero, the weaker the relationship.
How to calculate the two most common correlation coefficients are shown here.
A. Pearson Product-Moment (Pearson r)
The Pearson product-moment correlation coefficient is the most widely used measure of
relationship. The Pearson r is calculated to show the linear relationship between two variables. To
compute the Pearson r, two measures on each subject are needed. Suppose, for example, we have a
group of ten subjects, and for each subject we have measures of self-concept and achievement. We can
calculate the Pearson r between self-concept and achievement for these ten subjects, using the
following formula:

N ∑ XY −( ∑ X ) (∑ Y )
Pearson r =
√ N ∑ X 2−( ∑ X )2 • √ N ∑ Y 2−(∑ Y )2
 Where ∑XY is the sum of the XY cross products,
∑X is the sum of the X scores.
∑Y is the sum of Y scores.
∑X2 is the sum of the squared X scores.
∑ Y2 is the sum of the squared Y scores, and
N is the number of pairs of scores.

This formula may appear complex but is actually easy to calculate. The scores can be
listed in a table, as shown below; to use it, one simply finds the values for each summation formula,
substitutes where appropriate, and performs the math indicated.

Self-Concept Achievement
Score Score
Subject X X
2
Y Y
2
X •Y
1 25 625 85 7225 2125
2 20 400 90 8100 1800
3 21 441 80 6400 1680
4 18 324 70 4900 1260
5 15 225 75 5625 1125
6 17 289 80 6400 1360
7 14 196 75 5625 1050
8 15 225 70 4900 1050
9 12 144 75 5625 900
10 13 169 60 3600 780
2 ∑Y = 760 2 ∑X•Y = 13130
∑x = 170 ∑ x = 3038 ∑ Y = 3038
2
2
(∑ x ) = 28900 (∑ Y ) =577600

Step 1: Pair each set of scores; one set becomes X, the other Y
Step 2: Calculate ∑X and ∑Y
Step 3: Calculate x 2 and Y 2
 Step 4: Calculate ∑ x 2 and ∑ Y 2
Step 5: Calculate (∑ x )2 and (∑ Y )2

Step 6: Calculate X•Y

Step 7: Calculate ∑X•Y
Step 8: Substitute calculated values into formula

10 (13130 )−(170)(760)
Pearson r =
√10 ( 3038 )−28900 • √ 10 (58400 )−577600
131300−129200
=
√30380−28900 • √ 584000−577600
2100
=
√1480 • √6400
2100
=
( 38.47 ) •(80)

2100
=
3078

= 0.68

The value of 0.68 shows a moderate positive relationship between self-concept and
achievement for this set of scores.

B. Spearman Rank (r ranks or Spearman rho)

The Spearman rho is used when ranks are available on each of two variables for all subjects.
Ranks are simply listings of scores from highest to lowest. The Spearman rho correlation shows the
degree to which subjects maintain the same relative position on two measures. In other words, the
Spearman rho indicates how much agreement there is between the ranks of each variable.
The calculation of the Spearman ranks is simpler than calculating the Pearson r. The necessary
steps are

Step 1: Tank the Xs and Ys

 Step 2: Pair the ranked Xs and Ys.
Step 3: Calculate the difference in ranks for each pair.
Step 4: Square each difference.
Step 5: Sum the squared differences.
Step 6: Substitute calculated values into formula

The formula is
2
6∑D
Spearman rho = 2
n(n −1)

For the data used in calculating the Pearson r, the Spearman rho would be found as follows:

Self-Concept Achievement
Score Score
Subject X Y D D
2

1 1 2 −1 1
2 3 1 2 4
3 2 3.5 −1.5 2.25
4 4 5.5 −1.5 2.25
5 6.5 8 −1.5 2.25
6 5 3.5 1.5 2.25
7 8 8 0 0
8 6.5 5.5 1 1
9 10 8 2 4
10 9 10 −1 1

∑ D 2 = 20

Note: When ties in the ranking occur, all scores that are tied receive the average of the ranks involved.

6(20)
r ranks = 1−
10(100−1)

120
= 1− ¿
900 ¿
 = 1−0.12

= 0.88

In most data sets with more than 50 subjects, the Pearson r and Spearman rank will give almost
identical correlations. In the example illustrated here the Spearman is higher because of the low n and
the manner in which the ties in rankings resulted low difference scores.

Interpretation of Correlation Coefficients when Testing Research Hypothesis (Fraenkel, Wallen &
Hyun, 2013)

Magnitude of r Interpretation
.00 -- .40 Of little practical importance except in unusual circumstances;
Perhaps of theoretical value
.41 -- .60 Large enough to be of practical as well as theoretical value
.61 -- .80 Very important, but rarely obtained in educational research
.81 or above Possibly an error in calculation; if not, a very sizable relationship.

@@@
/Frajenkin-11-09-21
S