Sunday, September 16, 2012

Day 9_TEAM_C (Akash Singh)


Day 9_TEAM_C_Descriptive Analysis
-By Akash Singh

Discriminant Analysis?
Discriminant Analysis may be used for two objectives:
-Assess the adequacy of classification, given the group memberships of the objects under study
-Assigning objects to one of a number of (known) groups of objects.
It is an imperative tool in SPSS that helps in the prediction of the outcome. It can thus have a descriptive or a predictive objective.
Each method has a different approach; however the basic framework of all the techniques is the same:
In the given flow:
    Selection of panel members
    Term generation (or selection of appropriate lexicon)
    Concept formation
    Testing of panel consonance
    Evaluation of products

PANEL SELECTION:
The selection of panel members is very important to the quality of the data obtained. Potential members need to be screened for their ability to discriminate between similar samples, rate products for intensity and identify tastes and aromas. Equally, or possibly more, important than a panelists’ sensory acuity is their motivation. A panelist who feels they are required to participate may not perform as well as and equally skilled panelist who feels motivated to participate.

PANEL TRAINING:
Panel training encompasses term generation, concept alignment and panel testing phases. The amount of training required is dependent upon the method used as well as the product(s) to be tested. A company with an in-house descriptive panel may spend several months or more training a panel over a wide range of products, rather than training the panel specifically for each product as needed.

DA involves the determination of a linear equation like regression that will predict which group the case belongs to. The form of the equation or function is:

D = C + nX1 + nX2 + nX3+….niXi

Where, D = Discriminant function
             n = Discriminant coefficient
             X = Respondent’s score for that variable
             C= Constant
             i = the number of predictor variables

What is the purpose of Descriptive Analysis?

           Predictive DA addresses the question of how to assign new cases to groups.
          To determine the most thrifty way to distinguish between groups.
          To classify cases into groups. Statistical significance tests using chi-square enable you see how well the function separates the groups.
          To test theory whether cases are classified as predicted.

Submitted by: Team C
Akash Singh




No comments:

Post a Comment