 students now we are going to talk about factor analysis factor analysis actually is a method of scale validation when you develop a measure in order to understand any behavior, any concept for the empirical measure developed, we validate it so the procedure of its validation is statistically factor analysis factor analysis is considered a good tool to reduce the number of items in scale i.e. we can reduce dimensions too and we get to know that the available pool of items which of the items is best suited in your scale and the more variance you are explaining of your concept or the more coherent it is with each other so if we talk about factor analysis, it has two types of explorative factor analysis which is called EFA and confirmatory factor analysis which is called CFA so if we talk about EFA, EFA is generally used to discover the factor structure of a measure and to examine its internal reliability we theoretically developed a measure for that measure we made a set of items now we don't know which items are best suited to explain the concept and which items are best suited to the final measure so for that we need to do exploratory factor analysis so what happens with this is that in exploratory factors you don't go with fixed number of factors in fact, since we are exploring it now so that is why you see from the basis of data that how many factors emerge in one measure and what is the benefit of this is that when you are doing EFA you give all the chances to the items that all the factors are loaded if we talk about CFA, confirmatory factor analysis then this is A priori design in this we know that in our measure how many factors are there and how many items are loaded on one factor and which items are loaded on which factor i.e. we have developed a measure we are revalidating it we are checking its predictive validity we are checking its criterion validity that how the results come in different situations so for that we do confirmatory factor analysis in which we know that one item is loaded on factor 1 or factor 3 and in which factor, for instance factor 1 how many items are loaded and which is loaded and in factor 2, how many items are loaded and which is loaded and in factor 3, what are the items and what are the values so for factor analysis as you know we do some assumptions for hard test so for factor analysis that your items there should be no outlier in their data there should be a sample size advocate often it is said that if the sample size is not correct then the results of factor analysis are not correct so if 300 plus 350 is your sample then this is an advocate sample size there should be no multi-colonarity perfect there should not be any items in which there is absolute relationship between items the strong correlation is ideal that it should be kept in the same scale and there should not be any synacity as we had read in regression so here also there should be linearity there should be linear relationships and all the items should be made as continuous variable that is if there is a nominal item or an ordinal item then we do not add it in factor analysis and all the items should be in the same direction or in positive direction or in negative direction if there is any item in a negative direction like it is in bias then we do reverse coding for the factor analysis so I have given you a brief introduction of factor analysis in the next exercises we will talk in detail