I started doing a quantitative research for my thesis without any previous SPSS or proper statistic experience, and the first thing my professor told us in the spss workshop is that you start by measuring reliability of your data by testing for the Cronbach's alpha coefficient.
But when I started reading about it on my own, I found out that the Cronbach's alpha "is used to determine how much the items on a scale are measuring the same underlying dimension" and that it is very commonly used in psychological research where a questionnaire collects multiple measures of the same construct (like optimism or so).
But my study is different - I am working with a database that has information about companies and their patenting behavior (the number of patents they have, which industry those patents belong to, number of attorneys, inventors etc.) - to see if that information (independent variables) can predict future patenting behavior of those companies (dependent variable). So my independent variables do not measure the same construct (inventors are not the same as attorneys, for example) and I am wondering if testing reliability through Cronbach's alpha should be done in my case too. Maybe for my case there is some other way of measuring reliability?
I have tested the alpha it and it is 0.564, while the standardized alpha is 0.804. All of my variables are numerical but are not coded (numbers in some variables range from 0 to couple of thousands), so I believe coding them to something more standardized (like 0,1,2,3) will give better results (what are your thoughts on this?). But before I start doing that I wanted to see if I need to test the alpha at all in my case.
Thank you in advance and let me know if you have any questions.