Introduction to Bar Charts

[oldoc63]

Bar charts are generally used to visualize the relative frequencies of categories within a variable. Therefore, they are primarily helpful for visually summarizing categorical variables rather than quantitative variables. If you can organize your variable into distinct categories, a bar chart is often a great option.

Once you have your distinct categories, a bar chart is best used to display the different value counts of each category. We also compare means, but we would recommend using a side by side box plot for that instead because they give a full picture of the five-number summary.

[oldoc63]

Vertical vs. Horizontal Bar Charts

Within a bar chart, we have two axes. One axis represents the discrete values within a categorical variable we are comparing, while the other axis measures the counts of each of these values. Depending of how we orient these axes, we can either have a horizontal bar chart or a vertical bar chart.

[oldoc63]

You may notice that bar charts and histograms look almost identical. However, there are key differences between them:

• Bar charts are use for categorical variables, while histograms are used for quantitative data.
• Histograms must always be arranged in a specific order because they represent a range of numerical values. For a bar chart, each bar represent frequencies of category variables within a category. Unless the variable is ordinal, the bar could be arranged in any order.

https://static-assets.codecademy.com/Paths/data-analyst-career-path/barcharts_piecharts_lesson/bar_hist.svg

[oldoc63]

Bar Chart Area

Let's go over one of the most important concepts about bar charts. The bars of a chart have a couple of key features:

• They have lengths that are proportional to the counts they represent
• Each bar width in the bar chart is the same, meaning each bar's area is also proportional to the counts they represent.

These features make a bar chart super dependable for representing categorical data. For any chart like a bar chart, the areas we use to represent values must always be equivalent to the relative sizes of the values they represent. Otherwise, readers could be misled and potentially identify patterns that do not actually exist.