At some point in your life, I am sure you have taken a foreign language course. Any course in languages seems to linger in your mind. At the very least, you can recognize when a person is speaking the language because you remember the basics. Pronunciation of the alphabet comes to mind whenever I reminisce about my first Spanish language course. The more classes you take the less effort it is to recall the fundamentals. This also applies to statistical programming. When you learn a computer language, the objectives translate to other computer languages. Statistical programming is no different from learning a new language. There are rules. If you violate any of the rules, the computer will not understand you. You must use the correct syntax and “sentence structure” to execute your analysis. Personally, I have come across so many variations of a statistical programming language that I can simplify the objectives for you. Every statistics software package will have a command for reading your file into the software. SAS uses a libname reference. R uses a read reference. SPSS has a get file reference. All of these programs need to call the data into the software. Why? Because the software can not analyze something that is not in its ‘view’. Another common theme is the output command. The whole point of running a program is to create change or illustrate what you have so the output command is in every statistics software package.And of course you want to ensure your program is running so there is the log. The log is part of every statistics software package. The log has it’s own language and you must understand it in order to debug your program. That’s pretty much it. Everything else is a sub-command. Selecting cases, deleting records , calculating new variables, graphing, actual analysis- these are features which may vary with the software.
So the next time you learn a new statistics package please don’t let it intimidate you. It’s just another language.
If you enjoyed this blog there are more to follow-Amy