Once you've established what you'll be directly measuring you will then need to define which variables you will be using which are calculated based on your measurements.  Calculated variables can also be based on other calculated variables, but they all are eventually based on some set of direct measurements.

An example of a calculated variable could be population density by zip code.  This would take the measured population of a zip code, and then divide it by the measured area of that same zip code.

An example of a calculated variable which is derived from other calculated variables would be median national income on a purchase power parity basis.  Median incomes for a nation are calculated based on that nation's income distribution, and purchase power parity is calculated based on the cost of an average basket of goods for a given national currency.

At this point you should begin to define what your calculated variables are, and on what measured variables they're based.