Dear Shanjida Chowdhury ,

I don’t think you have done this correctly. Dividing by max - min has made your values all between 0 and 1. Z scores should be normally distributed around zero with a standard deviation of 1. If you wish to create a variable from a PCA I would recommend saving the regression scores with your components.

@ Peter Samuels thanks for your response. I am using regression score to create variable but i want to categorize that variable to low, medium and high categories....

To create the Z score, first find the Mean, SD of the values then using the formula

Z = (X - Mean)/SD, you get the Z score for each X value. This Z score has definite known properties while the scores you have created cannot be called as the Z score. In your Score, you are dividing each value (after taking deviations from the Minimum) by the range. It is better to follow Z score. Based on Z scores obtained you can create the three categories. I am not clear what do you mean by regression score. Are you using some model to arrive at the scores???

Ramnath Takiar Sir, Thanks for your response. From the factor score, I am creating a variable, not the regression score. First, I am performing PCA. Then, using eigenvalue as weight and factor score, I have created one variable. That variable has both negative and positive values. So you recommend using the z score formula rather than the [(Value-min)/(max-min)] formula.Now my question is, do I need to meet any criteria to use this formula that you said. After that, I need to create three categories named "low," "moderate, and "high." Is there any procedure to categorize that variable?

Unless I’m misunderstanding something z scores aren’t really part of a factor analysis except to show your confidence in the loadings. It’s the loadings that matter in an EFA.

Anuraj Nayarisseri yes