Although researchers refer to variables by a few different names, they can all be categorized as either quantitative or categorical. Knowing the difference allows the researcher to determine what kind of statistics to use and what inferences to draw based on the type of variable in a research project.
Anything researchers can count is a called a variable, including characteristics, such as color, number and quantity. All data collected in a research study regardless are whether you, as the experimenter, can manipulate it, is a variable. Examples include age, gender, income, country of birth and eye color. Researchers named these data "variables" because their value may change over time or between data units in a population. A variable, for example the number of compliments people give in an average week, varies depending on the type of people counted, such as boys versus girls or teenagers versus adults.
Nominal and Ordinal
Researchers classify numbers measured on a nominal or ordinal scale as categorical variables. Nominal variables use number as ways of naming data, rather than to count things. Examples would include 1 = yes and 2 = no. Ordinal variables use numbers to denote order, such as placing first or second in a race. Categorical variables measure the qualities or characteristics of data, so researchers call these qualitative variables.
Interval and Ratio
Researchers classify all numbers measured on an interval or ratio scale as quantitative variables. Researchers use interval scales when the difference between two values is constant. For example, 100 degrees is twice as hot as 50 degrees, which is twice as hot as 25 degrees. Ratio scales are interval scales with a true zero point. For example, 0 inches means that nothing is present so inches is a ratio scale; 0 degrees, however; does not mean that there is no temperature, so temperature is an interval scale. Researchers call quantitative variables that can have fractional values continuous, whereas they call quantitative variables that cannot have fractional values discrete.
Understanding what type of variable you have is the first step in deciding what type of statistical procedure you need to use. Other factors include the type of distribution, how many independent variables you have and what kind of dependent variables you have. For example, if you have one independent variable with two groups, a quantitative dependent variable and your data is normally distributed, you may want to apply a statistical test such as the t-test, or Fisher's exact test.