Discriminant analysis addresses the

situation of a non-metric dependent variable. In this type of situation, the

researcher is interested in the prediction explanation of the relationship that

affect the category in which an object is located, such as why a person is or

is not a customer, or if a firm will succeed or fail.

Formulation of discriminant analysis is:

Y1= X1 + X2 +…………. + Xn

Where,

Y1 –

dependent non-metric variable

X1, X2… Xn – Independent metric variables

The objectives for applying discriminant

analysis should further clarify its nature. It can address any of the following

research objectives:

Determining whether statistically

significant difference exist between the average score profiles on a set

of variables for two (or more) a priori defined groups.

Determining

which of the independent variables most account for the differences in the

average score profiles of the two or more group.

Assumptions:

1. The groups formed are mutually exclusive

and the group sizes are not different.

2. Independent variable’s variance

structure are same within each and every group of the

dependent variable.

Fallacies

are casually scattered.

The

purpose of the multi discriminant analysis is to explore differences among

groups, to eliminate the variables which are not or very less connected to

group variance, to categorize the cases into groups and to experiment theory by

watching whether cases are categorized as predicted. (George

H. Dunteman (1984).

Discriminant

analysis determines the optimal amalgamation of variables. Hence, the 1st

function gives the maximum discrimination between groups, the 2nd

gives the second most so on and so forth. If a more number of independent

variables are chosen and a functional subset has to be chosen for forecasting

the dependent variable, then multi discriminant analysis is used. (Morrison, D.F. 1967.)