Day 8- Team F(parveen)

Day 8

our learning during the day was factor Analysis

Definition of Factor Analysis

Factor analysis is a method of data reduction. It does this by seeking underlying unobservable (latent) variables that are reflected in the observed variables (manifest variables). There are many different methods that can be used to conduct a factor analysis (such as principal axis factor, maximum likelihood, generalized least squares, unweighted least squares), There are also many different types of rotations that can be done after the initial extraction of factors, including orthogonal rotations, such as varimax and equimax, which impose the restriction that the factors cannot be correlated, and oblique rotations, such as promax, which allow the factors to be correlated with one another. Factor analysis is a technique that requires a large sample size. Factor analysis is based on the correlation matrix of the variables involved, and correlations usually need a large sample size before they stabilize

Point of consideration during  Factor Analysis

1) Reduce no. of variable – So find what all variables are  related the each other and form the group

2) Find weather there is a common underlying factor between them ( eg:- we have a group of variable which are co-related from 14 we  can reduce to 2-3 and then we can get a better analysis)

How to do factor analysis

Analysis --à Data Reduction

Point to remember while doing factor analysis:-

1.       cant consider string model, we cant take binary or nominal ,

2.       only we had to consider scale variable

3.       In descriptive option – make sure the initial solution box is checked

4.       in extraction option – check the screen plot option and make the eigen value over= 1

5.       In Retation  = check the option varimax

Through this we had to analyze Variance of the value from the mean value of the variable

Variance = How much away the value is away from the mean

-         

      As all the variables are in different units , so during factor analysis we try to bring all variables on same level  by converting values to Z- scrore

Variance= No. value – average

Z score= variation/ std deviation

Zscore is used as normalizing method

Zscore has mean zero  & variance equal to 1

n  To analyze in a better way we can plot that analysis in graph- scatter-simple scatter

n  We can analyze Zscore value and variable value graph is same ,after reducing the horizontal axis cross value in Zscore graph to –ve value..

n  The main use is when we over lap the Zscore ,we can find the common extraction ,

n  Extraction values derived from the SPSS show what is the common of that variable with all other variable

- Parveen Rathee

Team F