Types of Numeric Data

There are three different types of numeric data:

• Nominal
• Ordinal
• Scale

It is important to recognize which type you are dealing with because the appropriate mode of analysis is dependent on the type of data.

Nominal data is "named" data

• The numbers assigned do not have any real numerical value
  • Examples
    • male = 1, female = 2
    • member of the treatment group = 1, member of the control group = 2
• Sometimes when you get statistical information out of a table you will have to assign numeric value to categories in order to input it into statistical analysis software (e.g. SPSS, SAS)



• The central tendency of nominal data is measured by the mode

Ordinal data is ranked data

• The higher numbers have more of a particular variable than lower numbers, but the distance between the numbers does not represent an equal amount of the variable
  • Examples
    • Letter Grades (A-F)
    • Likert Scales: "strongly agree" represents more agreement than "somewhat agree," but the amount of agreement between "neither agree nor disagree" and "somewhat agree" may not be equal to the amount of agreement between "somewhat agree" and "strongly agree."
• The central tendency of ordinal data is measured by the median

Scale data is interval data

• The higher numbers not only have more of a particular variable than the lower numbers, but the distance between the numbers does represents equal amounts of the variable
  • Examples
    • Temperature
    • Scores on scaled tests, like IQ tests
    • GDP (Gross Domestic Product)
• The central tendency of scaled data is USUALLY measured by the mean
• In the case of skewed data, provide both the median and mean

Next