The development of student use of graphs and tables is a skill that takes practice, and many students do not have the opportunity to carefully consider their appropriate use.
January is the time in my class when students working on independent research projects are finalizing their data and preparing it for their papers and displays. Our class spends a day discussing how data should be displayed, before submitting their work for teacher/mentor and peer review the following week.
I share with students an excerpt from Robert Day’s How to Write and Publish a Scientific Paper. My copy was given to me by my father, who was an engineer, and was published in 1979 – so it has some delightful comments about now outdated practices in publishing. Updated copies are now available (I wish I could comment on the updates, but I haven’t seen it in person), or there are free pdfs of it available online.
- Its not a good idea, or good science, or good writing to put ALL your data into your paper just because you have it. Keep a log book, and if you want to have your data presentable and cleaned up (or if you did a lot of calculating in a spreadsheet program), then type/print it and put it in a data binder.
- Tables should display variables among different samples/specimens/conditions. Tables should not show all the things that the samples have in common.
- Tables in the results section should not describe the controls of the experiment (e.g., temperatures, number of replicate samples, etc).
- Any table with a large number of zeros or 100% should be critically evaluated. Can you state this more simply in words?
- Make sure your values are significant! Why use space to show data that doesn’t really tell you anything?
- Do not use more significant figures than necessary.
- Don’t include columns with no significant variation.
- Don’t include a word list. Just use the words.
- Choose the best way to show your data (table, graph) and only display it once. Choose which based on what you want to show – are the values important, or just the trend/relationship?
Making a Table
- Data should be organized so that like things read down in a column, not across rows.
- The goal is to make your information easier for your reader (or teacher or judges) to understand.
- Do not include details in the results tables of experimental details that should be in the materials and methods/experimental design section of your paper. (Your results section should have results, not experimental conditions).
- Make sure your table has a meaningful caption.
- Graphs are pictures of your data. These are great for showing relationships and trends.
- Experienced audiences will notice when you are showing graphs of things that should be tables, or just a statement.
- Sometimes, the choice between table or graph is simply the choice of the author.
- If the data shows an interesting trend or picture, use a graph. It is much easier for the audience to grasp. Graphs provide clarity of expression that may not come from a table.
- Make sure to consider readability – don’t choose fancy fonts. sans serif fonts (those without serifs) are particularly easy to read.
- In a paper, choose marker points or lines that vary so your paper is readable in black and white. Don’t show different series of data with only variation in color.
- Make sure graphs are large enough to read.
- If there is space in the area of the graph, provide a legend or key.
- Crop photos to frame what you want your reader to see.
- Include a scale bar
- Write a good descriptive caption.