Statistics for everyone: ch.27 Factor analysis and cluster analysis

easier R than SPSS with Rcmdr : Contents

ch.27 Factor analysis and cluster analysis

Let’s explain a very rough concept. You are the teacher who now has your studens’ scores in Excel. Imagine that the names of the students are listed in vertical lines in column 1 , and the names of each subject are listed in row 1.

You can think of 2 classification systems: The first is the classification of subjects. It can be classified into a variety of subjects: Literature, Science, Arts, Mathematics, History, and so on. Each person will classify it differently, but they will classify it with their own criteria. It can be divided into subjects you like and dislike , or you can divide it into common subjects and electives.

The second is the classification of students. They can also be divided into the higher class and the lower class. You can also categorize them in the same way as math classes, science classes, theater classes, choir classes, etc.

In this way, classifying subjects corresponds to factor analysis, and classifying students corresponds to cluster analysis. Just from a table in Excel , it seems that it doesn’t matter if the horizontal and vertical changes from one to the other, but the information in this landscape and portrait is very different.

Columns, or vertical lines, are related to dimensions as subjects. 2 means 2-D, and 3 means 3-D information. Think of the previous drawing of a 2D and a 3D scatterplot. To have one more dimension is to provide a huge amount of information. If you take it as a health diagnosis, the weight and height information are 2D information, and when you add age to it, it becomes 3D information. If a teacher measure your language, English, and math scores, you data are 3D information.

A single horizontal line, is of 1 more student. Although 20 people are more information than 10 people, that is not a dimensional change. It can be difficult to understand these differences accurately. But I use this example for lectures .

Next Part is for Samples Size Calculation