In any essay or academic paper with a quantitative component, you’ll have to describe your variables at some point. There are four types of variables, each of which we’ve explained below:
Nominal Variables
Because nominal variables are qualitative in nature, they don’t fit any intrinsic numeric logic. You can’t measure the distance between the values within a numeric variable. For example, in measuring marital status, the possibilities might be (a) married, (b) single, or (c) separated. In entering nominal data into a statistical program, you could use any coding scheme you wanted, for instance:
1 = married
2 = single
3 = separated
Or, alternatively,
1 = single
2 = separated
3 = married
You’ll want to numerically code numerical variables so that you can easily create subsets in your subsequent statistical analysis.
Ordinal Variables
Ordinal variables represent progressive stages. For example, Olympic medals are ordinally measured: Silver is greater than bronze, and gold is greater than silver. However, as far as ordinal variables are concerned, there’s no precise measurement of the distance between levels. How much more important is a silver medal than a bronze medal? How much more advanced is stage IV vs. stage III cancer? Ordinal variables sort values into an order, but we can’t precisely specify the distance between these values.
Ratio Variables
A ratio variable has two properties. First, the distance between ratio values is precisely measurable. Second, there is an absolute, non-arbitrary 0. The distance between a pulse of 96 and 98 is precisely measurable (it is 2). Also, if you were dead, your pulse would be exactly 0. So, because the distance between pulse values is precisely measurable, and because there is an absolute, non-arbitrary 0 value possible for a pulse, this variable is a ratio variable.
Interval Variables
An interval variable is a ratio variable without an absolute, non-arbitrary 0. The Celsius temperature scale is interval because (a) the distance between Celsius temperatures is measurable but (b) the Celsius value of 0 is arbitrary. A Celsius value of 0 is the freezing point of water; it is not the absence of temperature! For example, a pulse of 0 is the absence of a pulse, which makes pulse a ratio rather than an interval variable.
Conclusion
Understanding the types of variables you’re working with is absolutely necessary for statistical reporting purposes in theses and academic papers. For example, when working with nominal and ordinal variables, you can’t measure the central tendencies (such as the means and standard deviations) of the variables. On the other hand, when you’re working with ratio and interval variables, you can—and should—both measure and report on various measures of central tendency, including mean, standard deviation, skewness, kurtosis, median, interquartile range, 95% confidence interval, and mode.
BridgeText can help you with all of your statistical analysis needs.
