Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

Statistical Tests applicable to Aquaculture Research, by RN Subudhi, Book-Chapter, in

“Applied Bioinformatics, Statistics & Economics in Fisheries Research”, Edited by AK Roy & N.
Sarangi, published by New India Publishing Agency, New Delhi [2008] ISBN 10: 81-89422-86-3

Details about the book from the publisher’s web: http://www.bookfactoryindia.com/index.php


Mail: newindiapublishingagency@gmail.com

Brief outline of the article:


The article tries to summarise the various statistical tests which can usually be used Fisheries Sciences
including Aquaculture Research. It first presents the formal research process and the different stages/
steps of (experimental/ empirical) research, so that the researcher knows where and why a statistical test
may be used.

It presents the formula used for small sample and large sample tests, and also the Non-parametric Tests
(NPT). It discusses analysis of nominal and ordinal data.

In Fisheries Sciences, researchers often handle quantitative data and cross-classified/ tabulated
records for analysis. In some cases quantitative measurement is not appropriate or possible. In the
investigation, we then make only the qualitative assessments. In such cases, we use two types of
qualitative variables:
a) Nominal Variables, and
b) Ordinal Variables.

The nominal or categorical data are usually classified according to some attributes or traits.
Category labels are just the names, and have no quantitative consideration. When put it in a cross
classification table, our objective would then usually be “to study the existence of possible association
among the categories”. The entries in the various cells of the table are only the “number of observations”
of each combination.
Let us take the example of the following table:

Observation on Growth of Prawn: Is it associated/ dependent significantly on water-(body)-quality?

Growth level
Water base Very good Average Poor Row Total
Pond a b (0) a+b
River c d (0) c+d
Sea e f g e+f+g
Column Total a+c+e b+d+f g Grand Total

In the above table (with observed data), we may be interested to know, if there is valid or really
significant association (dependence) between the growth of fish (dependent variable) with water-base
type (independent variable). This type of table can be generalized, and we may expect much more
combinations to occur in real field study.
Yule’s coefficient of Association is widely used such cases. Yule’s coefficient of Association, Q, is
given by:
Q = [(AB). ( ) – ( A ).( B) ] / [ (AB). ( ) + ( A ).( B)]
……………………………………………………………………………………………………………….
For the complete article, please refer the Book, published by NIPA, details as given above.

You might also like