Basically, the purpose of factor analysis is:

1. Summarizing the data, namely identifying the relationship between variables by conducting a correlation test.

2. Data Reduction, ie after doing the correlation, then proceed with the process of creating a new set of variables that are factors to determine a certain number of variables.

evaluating companies based on five items can use factor analysis if you want to differentiate and then group priority items based on rankings.

Do you have any candidates for hidden variables? Do you have a path-diagram relationship in your mind?

Factor analysis is overall better. But, meeting the presumptions and rationale behind the model are needed before implementing it.

M. Ichsan Nawawi thank you for you advice, I don't think that in my case the factor analysis makes sense as I don't need to group items or create new variables. I simply want to make a variable called "perception" which will demonstrate how respondents perceive a company. However, I'm having a problems to justify or argument my idea (why in this case, i think it is better to perform a mean value).

Babak Jamshidi Thank you, i understand that some presumptions need to be met. Now, I haven't found any hidden variables or anything that would prevent me from makinf factor analysis. Just in my mind, I have a feeling that factor analysis doesn't make sense, as I don't need to create new variables or group them, I need a variable which will address how respondents perceive a company. My only problem is, that I can't justify my opinion or motivation why I think it is better option.

I understand your reason (lack of grouping). You can check the direct and indirect effects. However, this book is helpful for you:

Latent Variable Models

An Introduction to Factor, Path,

and Structural Equation Analysis

Fifth Edition

Loehlin and Beaujean

I think this great book enables you to justify your answer better.