What is a Frequency Table?
A frequency table lists a set of values and how often each one appears. Frequency is the number of times a specific data value occurs in your dataset. These tables help you understand which data values are common and which are rare. These tables organize your data and are an effective way to present the results to others. Frequency tables are also known as frequency distributions because they allow you to understand the distribution of values in your dataset.
For example, if 18 students have pet dogs, dog ownership has a frequency of 18. A frequency table of pet ownership will list various types of pets and their frequencies, including dogs.
Frequency distribution tables are a great way to find the mode for datasets.
In this post, learn how to create and interpret frequency tables for different types of data. I’ll also show you the next steps for a more thorough analysis.
How to Make Frequency Distribution Tables for Different Data Types
You can make frequency tables for various types of data, including categorical, ordinal, and continuous. Categorical and ordinal data have natural groupings that you’ll use in the frequency distribution. However, for continuous data, you need to create logical groups for the frequency distribution.
Frequency tables display distributions for one variable, such as type of pet or dining satisfaction. When you need to assess two categorical variables together, use a contingency table instead. Learn more about Contingency Table: Definition, Examples & Interpreting.
Let’s go through examples of frequency tables for different data types.
For categorical data, make a frequency table by counting the number of times each group appears in your dataset.
Imagine you survey a class and ask them to indicate the types of pets they have. Type of pet is a categorical variable. Your raw data might be a list like the following:
From the raw data, count the occurrence of each type of pet and record them in the table. Because the categories don’t have a natural order, you can choose the order to list them in the frequency distribution that makes the most sense for your project. One option is to list the groups from most to least common.
In the example, I list the categories in descending order of occurrence, placing the most popular pets are at the top.
The frequency table indicates that dogs are the most popular type of pet among class members. Fish are rare pets in this class. Ten individuals do not have any pets.
Ordinal variables have at least three categories that have a natural order. The groups are ranked, but the differences between them might not be equal. For example, first, second, and third in a race are ordinal data. Learn more about Ordinal Data: Definition, Examples & Analysis.
For ordinal data, make a frequency table by counting the number of times each category occurs in your dataset.
Suppose you survey diners at a restaurant and ask them to rate their dining experience on the following ordinal scale:
- Very satisfied
- Very dissatisfied
Your dataset might look like the following:
From the raw data, count the occurrence of each level of satisfaction and record them in the frequency table. Because the groups have a natural order, list them in the frequency table using that order. In the example, I list the categories in descending order of satisfaction.
The frequency table shows that, on the whole, most diners were very satisfied and satisfied with their experience. However, there were a few diners who were not happy.
Continuous variables can take on almost any value, and you can divide them meaningfully into smaller increments, such as decimal values. Typically, you’ll measure continuous data on a scale. For example, when you measure height, weight, and temperature, you have continuous data.
Continuous data requires you to create the groups for frequency tables because they can have many distinct values.
Imagine you’re creating a frequency table of heights for 88 participants in a study. Your data will likely have many unique values. Below is a portion of heights in meters from an actual study I conducted involving preteen girls:
If you don’t create groups for continuous data like the example above, your distribution will contain many rows, each with a low count. That’s not going to be very helpful!
To make frequency distribution for continuous data, you’ll need to create groups of values for your continuous data. You can base your groups on ranges of values that make sense for your data when that’s possible. Usually, the spread of values for each group should be equal. In the frequency table, list these groups in ascending order. Groups must be mutually exclusive so that each data point falls into only one group!
In a frequency table for continuous data, the group counts indicate the number of times data values fall within each group.
For the height data, the frequency table indicates that a plurality of values falls near the center of the distribution (1.46 – 1.51m, f = 31). As you move away from the center, the occurrences decrease. The groups with the shortest and tallest heights have the lowest counts, 4 and 6, respectively. You can also see that the overall sample of heights ranges from 1.34 to 1.69m.
Next Steps After Making a Frequency Table
Analysts often create graphs that visually represent a frequency distribution because it gives their report more visual impact. Just like how you alter the frequency tables by the type of data, you’ll need to use various kinds of charts for different data types. Learn more in my post about graphing different types of data.
Making a frequency table is only the first step in understanding the distribution of values in your dataset. To better understand your data’s distribution, consider the following steps:
- Find the cumulative frequency distribution.
- Create a relative frequency distribution.
- Find the central tendency of your data.
- Understand the variability of your data.
- Calculate the descriptive statistics for your sample.
- Identify the probability distribution that your data follow.