| Module 2: Describing, Clarifying and Presenting Data
3. Data Displays
3.2. The features of a good table [4]
The American Psychological Association and the Modern Languages Association provide detailed information on how to construct tables. However, writers of newspaper and magazine articles, advertising and other reports might not follow the same guidelines. Below are a few simple steps that can be followed to produce clear and concise tables and graphic displays.
Layout of the table
The data should be organised so that the meaning the author is trying to communicate is clear. Since people find it easier to make comparisons when scanning a column of numbers than when scanning a row, variables are usually placed in columns and the units in rows. Generally, it is of more interest to compare values of the same variable in different population units than to compare values of different variables for the same population unit.
Table 1 and Table 2 below show the responses collected from university students who were surveyed about where they obtained their lunch. In both tables the venues are listed alphabetically. Table 1 is much clearer than Table 2.
While Table 1 is clearer than Table 2, it could be made even clearer by ranking the rows (i.e. by putting them in order of increasing / decreasing size) instead of listing them alphabetically. By ranking the rows, the comparison of the popularity of the venues becomes more obvious and the table is more useful, as shown in Table 3.
Have the values in the table been rounded?
Normally, a table is used to aid communication, and rounding values can help the reader to note the patterns and exceptions in the data. However, rounding might not always be appropriate. The example above is not rounded because we are counting students.
Does the table make sense?
Good tables should supplement the text and should make sense without the reader needing to refer to the text.
Tables should therefore consist of:
- a title (to provide a context for the table and tell the reader clearly what is in the table);
- headings or labels (to establish the organisation of the data and the identity of the variables placed in the columns); and
- a key to identify the size of the measurements (if required).
The table below was developed from a report conducted by KPMG Consulting called the Population Growth Report. The report used data from the Australian Bureau of Statistics to examine population growth and decline across Australia. Look at the table and think about how it could be changed to make it easier to interpret.
In order to interpret it easily you need to ask at least two questions.
1. ‘What do the authors of this table want people to compare?’ and, therefore,
2. ‘What should be the focus of the table?’
Do the authors want you to look at the changes in the level of state migration between 1996-1997 and 1997-1998? If they want you to compare different states, then perhaps the states should be listed in the table from highest to lowest net interstate migration. Alternatively, the interests of the intended audience can influence the layout. A reporter writing for the Hobart Mercury might sensibly put Tasmania in the top row and then reorder the other states.
Various changes could be made to this table to make it more informative. You might have suggestions for improving the table that are additional to the two below.
Suggestion 1:
Rank the table information from the most positive to the most negative values of the 1996-1997 net migration variable.
Suggestion 2:
It would be easier to compare the net migration from each state by rounding the numbers. Note that further rounding to the nearest hundred or even to the nearest thousand could still preserve the important signals in the data.
| 
