📂Data Science

Bar Graphs of Qualitative Data

Definition ¹

Assume that the frequency distribution of qualitative data is given.

1. A graph that represents the frequency with the height of the bars is called a Bar Chart.
2. A graph that represents the relative frequency with the area of the sectors is called a Pie Chart.

Description

Differences between the Two Charts

Obviously, the shapes are dissimilar, but the purpose and advantages/disadvantages of the two graphs differ.

• Unlike the bar chart, the pie chart is extremely effective in representing percentages, that is, the proportion within the whole. This difference stems from its inherent shape; while a circle’s interior angle is precisely 360 degrees, bars do not have this restraint, making it challenging to visualize the ‘whole’ at one glance.
• The pie chart tends to rely on color or shape more heavily, whereas the bar chart can be illustrated clearly using just one color. Thus, to decorate or emphasize the graph, the bar chart is much more convenient.
• The bar chart facilitates easy comparisons of magnitudes or hierarchies by aligning horizontal lines.

It might sound like a naive question to ask which chart is superior, but generally, the bar graph is better. As mentioned, when the proportion within the whole is extreme, that is, when the intention is not ’to deliver detailed and objective information’ but rather to assert that ‘a certain category constitutes the majority/minority,’ the pie chart has an advantage. In most other cases, the bar graph fares better.

Nightingale’s Rose Diagram

Florence Nightingale, a British nurse, associated the causes of death in the British army during the war with poor ‘sanitation’ of the time. The challenge was explaining this to the high-ranking officials in an understandable manner, considering ① the common knowledge that bacterial infection is more lethal than bullets was absent at the time and ② the foundation of statistics was still lacking. Pioneering in both fields, she devised the Rose Diagram and persistently persuaded the British government.

The diagram categorizes the year into twelve months, filling the outermost blue with the number of deaths due to infectious diseases, the middle dark color with deaths due to fatal injuries, and the innermost red with deaths due to other causes. It would be foolish not to see from this diagram that infections were the most significant issue. ^{2 3}

Looking back, the Bar Chart and Pie Chart are mixed in form, viewing which actually led the British to improve public health and successfully reduce the mortality rate among the British army. For this achievement, Nightingale is recognized both as a pioneer in nursing and in [descriptive statistics], endlessly mentioned by modern scholars as ’the importance of conveying statistical analyses understandably to non-specialists.’

Beware: Three Types of Lies

There’s a saying left by Mark Twain, the author of ‘The Adventures of Tom Sawyer.’

There are three kinds of lies: lies, damned lies, and statistics.

Translated, it roughly means ‘There are three types of lies in the world: lies, damned lies, and statistics.’

Because bar charts and pie charts are graphs easily understood by non-specialists and the general public, historically, they have been used in numerous misleading propagandas. ⁴ The image above seems to show a difference in numbers through the height of fountain pens, resembling a bar graph at first glance. However, on closer inspection, both the height and width increase, so the final area taken up by the leftmost pen is significantly exaggerated compared to the actual number difference. Although numbers are not omitted, and heights are not distorted, the general public, without meticulous scrutiny, may unknowingly accept some malicious tampering.

Let’s not do this:

• Choosing farming over employment… Young rural migrants increase by 50% in 6 years / KBS 2023.05.29.: https://youtu.be/sTLCzTEaL0s?t=41

See Also

• Histogram: A bar graph for quantitative data.