5.2 Describing Categorical Variables
describe groups or categories. Examples include gender, class level, major, condition, or whether someone completed a task.
For categorical variables, the main descriptive statistics are:
- counts, often written as
n - percentages, often written as
% - sometimes the mode, or most common category
Means and standard deviations usually do not make sense for categorical variables, especially nominal variables.
Frequencies and Percentages
A frequency table shows how many observations fall into each category. Percentages show the same information relative to the full sample.
For example, a class-level variable might include:
| Class level | Count | Percentage |
|---|---|---|
| First-year | 8 | 20% |
| Sophomore | 10 | 25% |
| Junior | 12 | 30% |
| Senior | 10 | 25% |
This tells us both the raw number of students in each category and how much of the sample each category represents.
Running Categorical Descriptives in jamovi
To describe categorical variables in jamovi:
- Go to Analyses.
- Select Exploration.
- Select Descriptives.
- Move your categorical variable into the Variables box.
- Select Frequency tables.
- Under Statistics, keep only options that make sense, such as
NandMissing.
If jamovi leaves a cell blank for a statistic, it may be because the statistic does not make sense for that variable type.
Interpreting Categorical Descriptives
When interpreting a frequency table, focus on the pattern of categories.
Ask yourself:
- Which category is most common?
- Which categories are rare?
- Are there unexpected categories?
- Are there duplicate categories caused by spelling, capitalization, or extra spaces?
- Are any categories so small that later group comparisons would be difficult?
This last point is important. If one category has only one or two people, descriptive statistics for that group may be unstable, and inferential comparisons may not be appropriate.
If a category appears more than once because of capitalization or spacing, clean the variable before reporting descriptives. For example, Female, female, and Female may look similar to us, but jamovi treats them as different categories.
A variable asks participants to choose their class level: first-year, sophomore, junior, or senior. What descriptive statistics should you use?
Answer
Use counts and percentages. Because the variable is categorical, frequencies are more appropriate than a mean or standard deviation.