Graphs and Charts and Tables, Oh My!

So you want to include a graph, chart, or table in your research report, but which one do you use? Is it just a matter of personal preference? Not exactly… The purpose of graphs, charts, and tables is to summarize information in an easy-to-understand way, but some types of data are better suited to graphs or charts, while more complicated or detailed information is better displayed in a table.

Graphs and charts are really the same thing, so we’ll treat them as one here, and are best used to illustrate a distribution of data.  But within the land of graphs there are several types, and the kind of graph you use will depend on what type of data you are reporting. For example, suppose you are an educational researcher and you want to display a grade distribution for a sample of students. Grades that are in numerical form are continuous data and so require a graph that represents this property such as a histogram or frequency polygon (see examples here http://onlinestatbook.com/2/graphing_distributions/graphing_distributions.html ). If instead, you want to represent the grade distribution in letter grades (A, B, C, etc.), it would be better to use a bar graph because of the categorical nature of this kind of data. You’ll notice that the bars on a bar graph are separate, as opposed to the bars that touch on a histogram, and this is reflective of the separateness of the categories (think back to the properties of ordinal or nominal data). Pie graphs, or pie charts, are another consideration and are a good way to represent proportional data. For example, if you want to show the percentage of students who earned each available grade (i.e. the percentage of students earning an “A”, etc.) a pie graph would show this very well.

What if the information you want to organize is more complicated than this? Tables are very handy for reporting things such as demographics of a sample, correlations between variables, or results of inferential analyses. A correlation table, for example, is a good way for your readers to see the relationships among several variables at one time. Then within the body of your paper you would list the most outstanding correlations and refer readers to the table to see the rest (see an example of a table here http://people.oregonstate.edu/~acock/tables/correlation.pdf ).

Graphs can be visually appealing, and most statistics programs will provide them, but its good to make sure you really need one before you put it into your research report. A good way to do this is to ask yourself if a graph will help organize your information, and if the answer is yes then you can decide which type of graph to use based on the nature of your data. Tables should also be used only when necessary, such as when you have a lot of information to present such as several correlations or the results of an inferential analysis. Always check with a formal source, such as APA to be sure you are forming your table according to standards. A popular resource can be seen here http://www.apastyle.org , and here https://owl.english.purdue.edu/owl/resource/560/19/ . At any rate, graphs and tables can help make your report professional and informative when chosen with necessity and purpose in mind.

Comments are closed