Although no one single visual display is most effective for presenting quantitative data, tables are often an ideal choice when you need to present specific values. Information placed within a grid framework and aesthetically designed for ease of use provides an efficient way for people to look up and compare data. Although we think of table data as typically numeric, values may also be presented as words.
Step inside the mind of someone reading a table and you’ll find there’s a lot of processing going on. So prior to embarking on the design, consider some of the mental tasks people might engage in when reading a table.
A person’s ability to quickly make use of table data is influenced by his or her familiarity with the tabular format, the complexity of the data and how well the table design matches its purpose.
According to a brilliant article from 1977, Rudiments of Numeracy, the author states that the criterion for a good table is that, “the patterns and exceptions should be obvious at a glance, at least once one knows what they are.” (Ehrenberg) The point being that reading a table is a repetitive task, and that an effective design allows people to clearly see the data as they become familiar with it. Understanding the data means a person is seeing how the numbers relate to each other.
Once you’ve identified the message that the data should communicate, the purpose of the table and the mental processes involved, consider which guidelines below will make your design most effective.
The more familiar users are with a table grid, the faster they can search it to extract the data. Meet the expectations of the audience by sticking to the conventions they expect, which are defined by the purpose of the table. For example, a table in a newspaper daily might include design flourishes one would never expect in a table for an annual report.
Structure and arrange the data to facilitate how it will be used. If the purpose of a table is to compare the population centers in a country, then organize the data from largest to smallest rather than alphabetically by city. If the purpose is to show the increase in college costs over a decade, then arrange the data by time.
To enable quick scanning, focus on the most important data and remove all extraneous information. Avoid clutter around the body of the table.
Use typography to create emphasis and to guide the reader’s eye. For example, headings can be larger or in bold type and highlights can provide emphasis. Visual cues can make a table easier to read so readers know what’s most important.
Populating a table with rounded integers makes it easier to read and to spot trends. Though this may not be appropriate for scientific data provided to experts, many tables are created for the general public who don’t need detailed accuracy. Consider how the numbers will be used to determine whether rounding makes sense. Statistician, Howard Wainer, recommends using no more than one decimal place in most cases.
When possible, don’t ask users to perform arithmetic computations or mental transformations with the data. Instead, do this for them by providing summary information in an additional row or column, such as total and averages. This facilitates quicker comprehension and interpretation.
When searching for information in a table, users expect the information to be displayed in a consistent manner. You can ensure there is consistency in the typeface of similar elements, in the alignment of similar data and in the emphasis of elements, such as column headings.
Intelligent use of alignment makes a table easier to read. Align all numbers, commas and decimal points with each other. Structure the table so it is clear that the data is aligned with headings and the grid in general.
To increase legibility, provide sufficient contrast between foreground and background. This can be an issue with table data if the rows or columns are shaded.
Due to the small size of visual memory and the difficulty of searching through complex information, reduce the number of columns when possible. If necessary, divide the table into two.
Side by side comparisons seem to be easier for people to make than above-below comparisons. In light of this, construct your tables so users will compare data between columns. In addition, the eye can run down a column rather quickly, but many people use their finger as a guide to read across rows.
If you can organize the data into subgroups and subcategories without altering the purpose of the table, this can improve search and make it easy to compare similar data.
Use the grid to guide the eye in the appropriate direction and to improve legibility. For an unobtrusive look, hide grid lines or display them as a subtle element. On the other hand, use strong grid lines when the information is complex. To draw the eye across the row, avoid vertical column lines or keep them subdued and use alternate bands of quiet color across rows to improve legibility.
Consider highlighting specific values to emphasize your message by drawing a box around the data or highlighting in a contrasting color.
If a subject matter expert is handy (or if you are the expert), provide one or two sentences to explain the main message of the table. This will hasten the reader’s understanding of the data.
Use white space between rows and columns, around headings, titles, labels and explanations. White space makes a table easier to read.
Labels are your opportunity to improve comprehension. Think through how the table will be used and the audience’s familiarity with the content and choose labels accordingly.
To avoid using little known abbreviations and acronyms that readers won’t understand, you may have to work your headings into submission. Consider spreading them over two or three lines or include heading detail in a footnote.
The lowly table is actually quite remarkable. It takes unwieldy data and compacts it into an organized structure. The designer then transforms this data into something meaningful that people can read and use. Why, it’s almost magical.