Frequency Tables

Frequency tables organize data by showing how often each value or category occurs. In brief: if the value 2 appears five times in a dataset, its frequency is 5. Dividing that count by the total number of observations gives its relative frequency.

A table makes repeated observations easier to summarize and prepares the data for graphs or further statistics. Its categories or intervals must be clearly defined so that every relevant observation is counted once. This guide explains simple and grouped tables, tallying, relative frequency and cumulative totals using a small hypothetical dataset.

What does frequency mean?

Frequency is the number of observations belonging to a specified category, value or interval. It is a count, so an ordinary unweighted frequency is a nonnegative integer. A frequency of zero means the category is included in the table but did not occur in the observed data.

The data value and its frequency have different roles. In a table of books read, a row labeled 3 with frequency 4 means four people each reported three books. It does not mean three people reported four books or that the total number of books is four.

A frequency table can summarize categorical data such as preferred transport or numerical data such as the number of siblings. The type of variable affects which later calculations are meaningful. Adding category codes as though they were measured numbers generally has no statistical interpretation.

How do you make a frequency table from raw data?

List the possible categories or distinct values, then go through the observations one at a time and tally the matching row. Grouping tally marks in fives can make counting easier on paper. After tallying, replace or accompany the marks with numerical frequencies.

Suppose a hypothetical class reports these numbers of books read: 0, 1, 2, 2, 3, 1, 4, 2, 0, 3, 2, 1, 4, 3, 2, 1. There are 16 observations. Counting gives frequencies 2, 4, 5, 3 and 2 for values 0 through 4 respectively.

Check the frequency sum against the number of raw observations: 2 + 4 + 5 + 3 + 2 = 16. If the totals disagree, an observation was omitted, duplicated or placed in more than one row. This total check is a basic safeguard before making a graph or calculating a statistic.

1. Identify the variable: number of books read.
2. List values 0, 1, 2, 3 and 4.
3. Tally each of the 16 observations exactly once.
4. Record frequencies 2, 4, 5, 3 and 2.
5. Verify that the frequencies sum to 16.

How do you calculate relative frequency?

Relative frequency is a category’s frequency divided by the total number of observations. For the value 2 in the example, the relative frequency is 5/16 = 0.3125. As a percentage, that is 31.25% of the class.

Relative frequencies allow comparisons between datasets of different sizes. A count of 5 in a group of 16 represents a different share from a count of 5 in a group of 50. The denominator identifies the group being described and must be kept consistent within the table.

Exact relative frequencies sum to 1, or percentages to 100%, when the categories form a complete nonoverlapping partition. Rounded displayed percentages may sum slightly above or below 100%. Do not alter an individual value arbitrarily without explaining a rounding adjustment; preserve the relationship with the underlying counts.

What is cumulative frequency?

Cumulative frequency adds frequencies in an ordered sequence. For numerical values listed from smallest to largest, it tells how many observations are at or below each value. In the example, the cumulative counts are 2, 6, 11, 14 and 16.

At the row for 2 books, a cumulative frequency of 11 means eleven people read at most two books. It is not the frequency of exactly two books, which is 5. The distinction between ‘exactly’ and ‘at most’ determines which column answers the question.

Cumulative frequency requires a meaningful order. Numerical values and ordered categories can support it, but an arbitrary list of favorite colors usually cannot. The final cumulative frequency should equal the total sample size, giving another consistency check on the table.

When should you use grouped frequency intervals?

Group numerical observations when there are too many distinct values for a readable row-by-row table. Heights, times and masses are common examples. Intervals compress the data but also hide the exact positions of individual observations within each group.

Define interval boundaries so that each observation belongs to exactly one group. A convention such as 10 ≤ x < 20, 20 ≤ x < 30 and 30 ≤ x < 40 assigns a boundary value of 20 to the second group. Writing simply 10–20 and 20–30 can be ambiguous unless the endpoint convention is stated.

Intervals should cover the intended range without unintended gaps or overlaps. Equal widths often make interpretation easier, but unequal widths can be used if clearly identified. A later histogram with unequal widths requires frequency density rather than treating bar heights as raw frequencies in the same way as equal-width groups.

What questions can a frequency table answer?

The largest frequency identifies the most common listed category or value. In the books-read example, 2 books is the mode because its frequency 5 exceeds the others. The table also shows that two people reported zero books and that five people reported either three or four books.

A numerical total can be reconstructed by multiplying each value by its frequency and adding: 0 × 2 + 1 × 4 + 2 × 5 + 3 × 3 + 4 × 2 = 31 books. The mean is therefore 31/16 = 1.9375 books per person.

For grouped data, exact numerical totals may no longer be recoverable because values within an interval are not shown. Using interval midpoints produces an estimate under that approximation, not an exact reconstruction. State whether the table preserves exact values or only grouped ranges before calculating further summaries.

How do missing data and multiple responses affect frequencies?

Decide how to handle missing observations before interpreting the total. If 20 people were asked but only 16 answered, a table of responses has denominator 16 unless the missing cases are represented separately. A percentage of respondents is not automatically a percentage of everyone asked.

For a question allowing multiple selections, category frequencies may sum to more than the number of people. Such a table is not a partition of respondents into exclusive groups. Label it clearly so that readers do not expect the percentages to total 100%.

The unit of observation also matters. A table might count people, transactions or repeated measurements from the same person. The frequency total tells how many observations were recorded, but it does not necessarily tell how many distinct individuals participated.

Why are frequency tables useful and which mistakes are common?

Frequency tables reduce a raw list to an organized count while preserving the features needed for many graphs and statistics. Their background is systematic classification: each observation is assigned according to a defined rule. Clear categories are therefore as important as correct addition.

Common mistakes include confusing values with frequencies, double-counting interval endpoints and using the wrong denominator for percentages. Another is interpreting grouped estimates as exact data or treating a multiple-response table as though categories were mutually exclusive.

Check the row definitions, the frequency total and any cumulative or relative-frequency columns before drawing conclusions. Keep the original data available when possible so discrepancies can be traced. A well-labeled table should let another reader determine exactly what was counted and what the totals represent.