You are on page 1of 3

MODULE 14

NON-PARAMETRIC STATISTICS: NATURE AND SPEARMAN RANK-ORDER CORRELATION

==================================================================================
OBJECTIVES: At the end of the lesson, the students are expected to:

1. Discuss the nature of non-parametric statistics.

2. Explain when to use Spearman Rank- Order correlation.

3. Solve statistical problems using Spearman Rank- Order correlation.

==================================================================================

Nature of Non-parametric Statistics

 Non-parametric test which is also called distribution-free statistics because it is free from
the assumption of a normal distribution in the population being studied.
 Non-parametric tests are used when dealing with nominal (counting) or ordinal data
(ranking).
 If the independent variables are non-metric, the non-parametric test is usually performed.
 Some non-parametric tests do not require a large sample.
 Some non-parametric tests are the Spearman Rank-Order Correlation, Kendal Tau Coefficient of
Concordance, the Chi-Square Test, the Kruskal Wallis-H Test and Mann Whitney-U test.

SPEARMAN RANK-ORDER CORRELATION ( SPEARMAN’S RHO)

 Non-parametric correlation statistic are used if the data are expressed in rank or data that are
converted into ranks (ordinal data).
 It is an alternative method to the Pearson Product-moment Correlation if the sample size is
small.

Formula : r = 1 – _6 (∑D2)__
N (N2 – 1)
Example:

Let X = GWA in College


Y = Annual Income in Thousand Php.

Respondent X Y Rx Ry D D2
A 86 287 2 1 1 1
B 79 138 7 6.5 .5 .25
C 84 200 4 4 0 0
D 91 245 1 2 1 1
E 82 200 5 4 1 1
F 77 125 8 8 0 0
G 80 138 6 6.5 .5 .25
H 85 200 3 4 1 1
∑= 4.5

Required: Determine if there is a significant correlation between GWA and Annual Income.
Use .01. Interpret result.

Steps:

1. State alternative hypotheses.

Ho: There is no significant correlation between GWA and annual income.


Ha: there is a significant correlation between GWA and annual income.

2. Rank the data. Follow rules in ranking.


3. Solve the correlation coefficient

rs = 1 – _6 (∑D2)__
N (N2 – 1)

rs = 1 – 6(4.5) = 1 – 27/8(63) = 1 – 27/504 = 1 – .054 = .95


8 (82 – 1)

4. Test the significance of the correlation.

n= 12 ; r.01= .678

rs= .95> r.01= .678: The computed correlation coefficient is significant; REJECT Ho.
5. Interpretation/Conclusion

There is a significant correlation between GWA and annual income. A very


high positive correlation is evident indicating a very strong positive relationship
between GWA and annual income.

You might also like