A case is the smallest entity that is being analyzed in a study. It is the rows of a raw data table. Often, but not always, each case is a person. Sometimes raw data files are inadvertently created with the wrong case.
Often, each case is a person
Often, each row in the raw data table contains the data for one person. For example:
- In most surveys, each case represents each person that answered a questionnaire. An example of such a data table is below.
- In studies of education, each case is often a student.
- In studies of poverty, each case is a person.
Common examples where each case is not a person
However, there are many, many, studies where each case is not a person. For example:
- With sales analysis, a case can represent many different things, such as :
- All the sales of a product at a point in time (e.g., the annual sales of iPhones), or,
- Each individual sale (e.g., each row of the data file may represent a separate iPhone).
- A survey of schools might have the school as the case, with one or more people in the school, providing information about the school.
- A study investigating how mice react to changes in diet may have the mice at different points in time as the case. For example, if there are 30 mice in the experiment, and the mice are observed prior to the change in diet, 1 week after the change in diet, and 2 weeks after the change in diet, then there are 90 cases.
The table below shows some examples of cases used in data analysis.
|Studies of organizations||Studies of people|
Number of employee grouping
Turnover grouping (e..g, under $1 million, etc.)
The choice of case is a decision
In most analyses, the choice of case is obvious and the issue is not really thought about. However, sometimes data sets are created with the wrong case, and the data needs to be reshaped to fix this problem. This is discussed in more detail in Reshaping a Data Set: Widening, Stacking, Aggregating, and Merging Data FIles.
Also known as
A case can also be referred to as some of the following, although in different contexts some of these can have different meanings:
- Unit of observation
- Unit of analysis
- Analysis unit
- Analysis subject
- Population unit
- Sampling unit
The next article examines the idea of a variable in more detail: Variable.