Categorical Data vs Numerical Data
Data are the facts or information collected for the purpose of reference or analysis. Often these data are collected as an attribute of the concerned subject. This attribute can vary from one to another hence this varying attribute can be considered as a variable. The variables can assume different forms of values and these are intrinsic in the collected data.
Variables can be either qualitative or quantitative; i.e. if the variable is quantitative, the answers are numbers and the magnitude of the attribute measured can be stated with a certain degree of accuracy. The other type, the qualitative variables measure the qualitative attributes and the values assumed by the variables cannot be given in terms of size or magnitude. The variables itself are known as categorical variables and the data collected by means of a categorical variable are categorical data.
More about Numerical Data
Numerical data are basically the quantitative data obtained from a variable, and the value has a sense of size/ magnitude. The Numerical data obtained are further divided into three more categories based on the theory developed by Stanley Smith Stevens. Numerical data can be either ordinal, interval or ratio. The type of the data is determined by the method of measurement of the values, and the types are known as levels of measurement.
The weight of a person, the distance between two points, temperature, and the price of a stock are examples of numerical data.
In statistics, majority of the methods is derived for the analysis of numerical data. Basic descriptive statistics and regression and other inferential methods are majorly used for analysis of numerical data.
More about Categorical Data
Categorical data are values for a qualitative variable, often a number, a word, or a symbol. They bring out the fact that the variable in the considered case belongs to one of the several choices available. Therefore, they belong to one of the categories; hence the name categorical.
The political affiliation of a person, nationality of a person, the favourite colour of a person, and the blood group of a patient are qualitative attributes. Sometimes, a number can be obtained as a categorical value, but the number itself does not represent the magnitude of the attribute measured. Postal code is one example.
Also, any categorical values belong to the nominal data type, which is another type based on the levels of measurements. Methods used for analysing categorical data are different from that of numerical data, but the underlying principle may be the same.
What is the difference between Categorical and Numerical Data?
• Numerical data are values obtained for quantitative variable, and carries a sense of magnitude related to the context of the variable (hence, they are always numbers or symbols carrying a numerical value). Categorical data are values obtained for a qualitative variable; categorical data numbers do not carry a sense of magnitude.
• Numerical data always belong to either ordinal, ratio, or interval type, whereas categorical data belong to nominal type.
• Methods used to analyse quantitative data are different from the methods used for categorical data, even if the principles are the same at least the application has significant differences.
• Numerical data are analysed using statistical methods in descriptive statistics, regression, time series and many more.
• For categorical data usually descriptive methods and graphical methods are employed. Some non-parametric tests are also used.