In multivariate analysis, one goal is to explain why observations, in this case people, have different answers on your survey. Factor analysis groups survey questions that vary together. This makes it easier to model in a regression or anova because it reduces a large number of variables (i.e., survey items) into a smaller set of factors. Each factor explains a percent of the total variance. Factors that do not explain much variance might not be worth including in the final model. It takes some iteration to come up with the optimal number of factors.

The canonical resource for factor analysis is Kim & Mueller (1978). UCLA has a great online resource.

http://www.ats.ucla.edu/stat/spss/output/factor1.htm

http://www.ats.ucla.edu/stat/seminars/muthen_08/part2.pdf

Many Thanks Professor Richard.

I also need clarification about

1. Will any dependent variable or anything else be added in "Selection Variable" when running Factor Analysis in SPSS (Picture FA Question.png)?

2. Can you provide little interpretation of picture FA TVE Table.png. I have found figures from tables (as in picture FA 2.png) for the 4 cases as

  i. Factor Analysis for items from 8 functions (called old Variables)

  ii. Factor Analysis for old (8) plus items of Function R.

  iii. Factor Analysis for old (8) plus items of Function O.

  iv. Factor Analysis for all 10 Function.

I am studying that inventory should include 10 factors that is shown by EFA and CFA but Total Variance explained is highest in Case ii above (Factor Analysis for old (8) plus items of Function R.).

Many Thanks

Pl see the link. May be of some use.

Chapter A note on the Factor Analysis in Forestry Sciences

I took a look at your output with regard to Explained variance, and it appears that the unrotated solution would support a single factor solution, since the amount of explained variance drops sharply after the first factor (which is known as a "scree" plot to assessing the of factors). 

If that is the case, then you will have relatively strong loadings (.4 or higher) on each of the variables for the first factor, and relatively small loadings on all the other factors.

It seems to me that you need to do more background reading on factor analysis, and here are links to some introductory sources:

http://pages.towson.edu/jgarczyn/scalesspss.pdf

http://ro.ecu.edu.au/cgi/viewcontent.cgi?article=1373&context=jephc

http://www.statisticshell.com/docs/factor.pdf

http://www.ucdenver.edu/academics/colleges/nursing/Documents/PDF/FactorAnalysisHowTo.pdf

I believe that all of these use SPSS, which was not intentional, it is just what showed up within the limits that I set.

Hi Mattiullah,

there are some things unclear for making interpretations. Taking explained variance has different meanings according to

a) do you conduct a principal component or a common factor analysis?

b) how much items are in your itemset?

a) In principal component analysis is used (only), in order to reduce the itemset from many to few. It bins variance of the items. This happens regardless from if there are any correlations among your variables. Thus, taken for itself, it is not possible to interpret. Correlation tables and an additional common factor analysis may be insightful. Both of the latter can indicate latent factors, which carry systematic variance regarding individuals differences. Thus when you use component analysis, go for as much explained variance as possible with a reasonable set of factors fewer in number than the complete set.

b) For simplification: In a set of 10 variables, 10% explained variance means that a "factor/component" can explain variance comparable to one variable... in a set of 100 variables 5% explained variance means that a "factor/component" can explain variance comparable to of 5 variables. This should be included in your decision.

Best, René 

In addition, the TVE must be significant, i.e the percentage.

what is meaning of variance extracted . pl explain in simple language

If total variance explained shows extraction can be done. What should I do?

Eigenvalue ≥1 , scree plot and the total amount of variability of the original variables explained by each factor solution are important. It is also important that the rotated factors make theoretical sense to the researcher.

We want to see in factor analysis a smaller number of interpretable factors that explain the maximum amount variability in the data. So , you are the best judge to consider to take factors according to your study needs.