Week 8
Week 8; Factor analysis
In this week we will discuss about the factor analysis on its meaning, objectives, types of factor analysis, and pros and cons of using factor analysis.
Factor Analysis
In this week we will discuss about the factor analysis on its meaning, objectives, types of factor analysis, and pros and cons of using factor analysis.
Factor Analysis
Factor analysis is a statistical technique to group or combine variables that are related both positive and negative in the same group. The central aim of factor analysis is the "orderly simplification" of several interrelated measures using mathematical procedures. Variables within the same factor is highly correlated while variable within the difference factor is less or no correlated. It can be used to develop a new theory or confirm the generalization.
Why use factor analysis
Factor analysis is a useful tool for all sciences that are concerned to discover if variables from regular patterns and vary together.
1. To investigate whether the common factor explained the relationship between variables, where the number of common factors is less than the number of variables. This model is known as the exploratory factor analysis model (EFA).
2. To test the hypothesis about the structure of factors, or whether this factor matches the model or the existing theory. This model is called Confirmatory factor analysis (CFA).
Types of factor analysis
1. Exploratory factor analysis (EFA) is used in case the researcher does not have knowledge or have less knowledge about the relationship structure of variables, to study the structure of variables and reduce the number of existing variables to be combined.
2. Confirmatory factor analysis (CFA) is used in case the researcher has known about the relationship structure of variables or expected how the relationship structure of the variable is and then use the confirmatory factor analysis to test or confirm the relationship as expected.
Advantages and disadvantages of Factor analysis
Advantages
1. Reduce the number of variables by combining several variables into the same factor. The factors are considered new variables that the value of the generated factor is called the Factor Score.
2. Solve the problem due to multicollinearity by combining the independent variables that are related.
3. Provide the relationship structure of the variables. The Factor Analysis technique will find the correlation coefficient of the variables, one by one. Then combine the corresponding variables into the same factor.
Disadvantages
1. Sample size, if the sample size is small, the correlation coefficient is low. Thus, the sample size must be large.
2. Data must be interval scale or ratio scale; the category data must be dummy variable and then be analysed. Moreover, the data must be normal distributed.
ความคิดเห็น
แสดงความคิดเห็น