Classifying the data and presenting it in tabular form makes the data compact and easy to grasp. To make the data even easier to understand and more visually appealing we express the data in the form of diagrams and graphs. There are many different types of diagrams and graphs that find a wide range of applications in statistics.
Some examples of graphs are bar graphs, line graphs, histograms, etc. Pie charts and pictograms are some examples of diagrams in statistics. We now list some of the advantages and disadvantages of using graphs and diagrams to present data in statistics.
Advantages of Diagrams and Graphs:
- A pictorial representation gives us an idea of the complete data at a single glance. It provides us with a birds-eye view of the data. This makes the data easy to comprehend.
- Diagrams are more attractive and appealing compared to rows and columns of numerical data.
- They are readily understandable even by laymen and those people who may not have statistical training or background.
- It takes a lot of effort to go through numerical data and draw meaningful inferences from it. On the other hand, graphs save both our time and effort since they are easily understood.
- Diagrams allow us to make comparisons of two or more data sets. For example, if we set pie charts showing the monthly breakdown of household expenses of two families side by side then it becomes very easy to do a comparative analysis.
- Time series graphs can help us detect trends in the data. Suppose we draw a line graph plotting the yearly profits of the company. If the line graph keeps trending upwards then we can conclude that the company is making greater and greater profits year after year.
Limitations of Diagrams and Graphs in Statistics:
- Diagrams and graphs tend to only show the broad and general aspects of the data. In order to understand the data with granular details, we need to classify and present the numerical data in tabular form.
- Though it is advisable to only use diagrams and graphs when presenting data to the general public and laymen, they are not enough for a professional statistician. A professional statistician always needs numerical data to carry out any analysis such as testing a hypothesis, carrying out an ANOVA test, etc.
- They are prone to misinterpretation and misuse. Diagrams can also be used for the purpose of false advertising. This can be done by changing the scale of the data, to make certain things appear bigger or larger. The general public might be deceived by dishonest individuals if they accept such graphs at face value.
- Many of the diagrams are not easy to construct. For example, drawing a three-dimensional bar graph takes a lot of time and effort.
- Some diagrams can easily get congested. For example, when constructing a pie chart if there are too many categories/slices then the resulting diagram becomes very ugly, congested, and difficult to read.
- If there are too many large values in the data, then drawing a graph might hide the very small but significant differences in data values.
- Each kind of data is best represented by a suitable kind of diagram. If the wrong type of diagram is used to present the data then it may lead to incorrect conclusions. For example, we cannot use a pie chart to represent non-categorical numerical data. We should instead use a histogram to present the data in this case. Pie charts are a suitable method to represent the data if there are 6 or fewer categories.
- Two different diagrams can be compared to each other only if both data sets are measured in the same unit. If the two data sets are in different units then we cannot compare the two diagrams.