Boxplot: Great for data analysis, not so much for presentations
The boxplot is a type of chart that is very useful for analyzing data distribution. With it, it’s possible to investigate:
- The spread of the dataset
- The symmetry of the dataset
- The distribution of the dataset
- Minimum and maximum values of the dataset
- Outliers or unusual values
In the image of this text, we have a boxplot created from fictitious values. In it, we can observe:
- Median (Q2): The central value of the dataset
- First quartile (Q1): The value that delimits the lower 25% of values, which normally coincides with the median between the Median and the minimum value (excluding outliers) of the dataset
- Third quartile (Q3): The value that delimits the upper 25% of values, which normally coincides with the median between the Median and the maximum value (excluding outliers) of the dataset
- Interquartile range (IQR): The range between Q1 and Q3, calculated by Q3 — Q1, which contains 50% of the dataset, i.e., half of its observations are within the “body” of the boxplot
- Outliers: Values considered atypical, normally represented by points beyond the maximum and minimum values
Furthermore, from the analysis of this visualization, we can also compare groups within the dataset we are studying to evaluate how information is distributed in each one, such as education, gender, and marital status.
Despite being a great chart for exploratory analysis, insight generation, and understanding the problem, the boxplot should be avoided in contexts of presenting results, especially to non-technical audiences. For this audience, the boxplot is not usually considered an intuitive or easy chart to interpret. When presenting results, the ideal is to extract the value from the boxplot and transform this information into another format, such as a bar chart or even text.