Data Types and Measurement Scales

Data can be separated into two main groups: It’s type and measurement level.

Data types can be categorical and numerical.

Categorical: It can be groups in things such as company brands, but also yes and no questions and these two answers would be classified into two categories.

Numerical: This is numerical data. This data can be separated into two groups:

  • Discrete: Discrete date is finite integers, like the number of children you’ll have. But physical money can also be discrete, since you can only pay one cent difference at a minimum. Similarly, time on a clock is discrete, time in general is continuous.
  • Continuous: pertains to a range of data that can vary endlessly, like your weight on a scale, this can vary almost infinitely. Height, area, distance and time can also be continuous, since they can vary in infinitely small amounts.

 

Measurement levels can be: Nominal, Ordinal, Interval and Ratio. These are the four variables you can capture with surveys. Basic summary: Measurement levels can be broken down into two groups: Qualitative and Quantitative data.

 

  • Nominal: These are names or or labels with no specific order, like “male and female,” “apple and orange.” There is no quantitative value or order. They can be the different brands of a company or seasons of the year.
  • Ordinal: This scale is used to determine a non-mathematical scale of something, such as satisfaction, happiness, degree of pain, etc.
  • Interval: Doesn’t have a true zero. Example is Celsius. If today it’s zero and tomorrow two degrees, it doesn’t mean it’s twice as warm.
  • Ratio: Have a true zero. Most things in the real world are ratios. If I have three apples, you have two, I have three times more apples since the ratio between six and two is three.

 

integers and decimals can both be intervals or ratios, but it depends on the contexts you’re working with.

Posted in Blog Posts, Computer/Data Science and Programming.