Recall the purpose of a frequency distribution is to summarize a set of data. We will fail at this purpose if the table contains too many categories. For this reason, continuous data are often organized into logical intervals and then listing the intervals in the table rather than each specific score. For example, a professor wanting to summarize students’ scores on a test could list each possible score from 0 to 100. This would likely produce a table larger than the actual column of raw scores. Since letter grades are assigned in ten point intervals, it would be logical to establish intervals that capture the lowest score and then proceed by increments of 10% until the highest score is captured. These groups of scores, or intervals, are often called class intervals.
If you are constructing a grouped frequency distribution table, examine the data as we did above to see if there is a logical number of categories to use. If no clear “natural” classification emerges, then construct your table with ten categories. Of course, you will want to adjust this number if ten categories result in an illogical presentation. Your ultimate goal is to present your data in a form that is easy to understand. It is best to make all of your intervals the same width. This advice is commonly ignored with a “catch all” category at the end of the list. Tables with much larger intervals in the final category can be misleading because that category because it is much broader than the other, may have the higher frequency. Consider this when you are the consumer of research.
The primary advantage of presenting data as a grouped frequency distribution is that it is easy to “see” the characteristics of the data at a glance. Remember, when you group data you are necessarily losing information.
[ Back | Contents | Next ]
Last Modified: 06/29/2018