You can think of a variable as a characteristic that varies across individual elements. For example, hair colour varies across individuals: black, blonde, brown, red, grey (or practically any colour if we include the wonders of hair dying). If we go by other physical characteristics, we can easily see that height, weight, body type, skin colour, age, etc. are all variables.

Then what about social/economic characteristics like level of education, annual income, occupation, employment, citizenship, marital status, political party affiliation, union membership, participation in sports (to name a few)…? All variables. Or, what about personal opinions and preferences? You might love chocolate a lot but your friend might not care for it; another friend might like it but just a little… Your friend might try to convince you that classical music is great but you might find it terribly boring, preferring rock instead. You might be a dog person and might frequently extol the virtues of dogs in comparison to cats, to the dismay of your cat-loving significant other. You might think that legalizing marijuana in Canada was the right decision but your parents might feel it was a profound mistake on part of the government. Clearly, opinions and preferences vary, so we can add ‘opinion on marijuana legalization’, ‘liking of chocolate’, ‘preferred music genre to listen to’, and ‘favourite pet animal‘ to our ever growing list of variables.

So far, you might decide that variables only apply to people: after all, all the examples mentioned above discuss characteristics that vary across human beings. However, this is absolutely not the case, as we can just as easily see that other things can have varying characteristics. For example, universities can differ in their student enrollment numbers, instructor-to-student ratios, type of degrees awarded, geographical location, source of funding, presence of medical school, percentage of international students, etc. Countries vary on population size, climate, geographical/geopolitical location, language, GDP (gross domestic product), level of human development, presence of minority groups, immigration (and emigration) rates, fertility and mortality rates, access to universal healthcare, average education level, age of majority, freedom of press, type of government… you get the picture. Clearly, variables apply to elements of anything that may be compared on characteristics which vary across these elements (hence the somewhat clumsy definition I started with).

Researchers refer to units of analysis when they want to specify the elements they study: When we have information about characteristics of people, we say that the unit of analysis is “individual”. When instead of people, we study countries, the unit of analysis is “country”, and so on.