What Is Relative Frequency?

Relative frequency is the proportion of times a value or category occurs in a dataset. If you have 50 students and 12 prefer pizza, the relative frequency of the pizza preference is 12 divided by 50, which equals 0.24 or 24%. The relative frequency is always a fraction of the total, never a raw count. The raw count, 12, is the frequency. The relative frequency is what that count means in relation to the whole group. This single idea underlies every frequency table, every bar chart, and every discussion of observed data in introductory statistics. The OpenStax 'Introductory Statistics 2e' text, in section 1.3, defines relative frequency as the fraction of times a value occurs, and that definition carries through all of chapter 2 on descriptive statistics. Without this distinction, the rest of statistics makes little sense, because every proportion, every percentage, and every probability you will meet later rests on the same logic of part over whole.

The relative frequency definition is precise: it is the frequency of a particular value divided by the total number of observations in the dataset. Written as a formula, it is f/n, where f is the count for that value and n is the total count. For example, if a die is rolled 60 times and the number 4 appears 9 times, the relative frequency of rolling a 4 is 9/60, which equals 0.15 or 15%. Notice what this does not say: it does not say anything about what the die will do on the next roll. It only describes what happened in those 60 rolls. This is the core of the relative frequency meaning in statistics: it is a description of observed data, not a prediction of future events. A value with a relative frequency close to 0 is rare in your sample; a value near 1 is very common. The OpenStax text, in section 1.3, makes this same point when it contrasts relative frequency with theoretical probability, and it returns to the idea in chapter 2 when building frequency tables. The definition never changes: relative frequency equals the count for a value divided by the total count, and it is always a number between 0 and 1 inclusive.

The relative frequency meaning goes beyond the arithmetic. It tells you how common a value is relative to the entire dataset, which is why the word 'relative' matters. A frequency of 20 means very little on its own; 20 out of 30 is a large share, while 20 out of 2000 is a small one. The relative frequency converts that raw count into a comparable proportion. This is what makes it useful for spotting patterns. If you survey 100 people and 80 say they prefer coffee, the relative frequency of coffee preference is 0.80, and you immediately know coffee dominates the group. If you only had the frequency, 80, you would still know the count but not its significance. The relative frequency meaning also includes the idea of the whole. Every relative frequency in a table is calculated against the same total, so the values are directly comparable across categories. […]

Frequency vs relative frequency is a distinction every student must master early. Consider a class of 20 students where 5 have blue eyes. The first is a whole number; the second is a proportion. The frequency tells you the actual number of observations, which is useful for exact counts. If another class has 10 students with blue eyes out of 40, the frequency is higher (10 vs 5) but the relative frequency is the same (0.25 vs 0.25). The raw count alone would mislead you into thinking the second class has a bigger blue-eye presence, when in fact the proportion is identical. This is the central lesson of the frequency vs relative frequency distinction: always ask which one you need before you interpret the number. […]

Relative frequency statistics is the term for using these proportions to summarize and describe data. It lets you compare a category with 40 out of 100 observations to a category with 400 out of 1000, even though the raw counts differ. The statistics themselves are simple arithmetic, but the interpretation is not trivial. The statistics also feed into cumulative relative frequency, which is the running total of relative frequencies and must end at 1.0 (or 100%) because it includes all values. These are three ways of saying the same thing, but they are not interchangeable in a table. A percent is often more intuitive for a general reader; a proportion is better for further calculation.

All relative frequencies in a table sum to 1.0 (or 100%) because every observation is counted in exactly one category. If you have 50 observations and you add up the relative frequencies of all categories, you are adding up all 50 observations, each expressed as a fraction of 50, so the total is 50/50, which equals 1. This is a logical necessity, not a convention. But here is where the rounding trap appears. If each relative frequency is rounded to two decimal places before summing, the total may be 0.99 or 1.01 instead of 1.00. This is not an error in your arithmetic; it is an error from rounding intermediate values. The correct practice is to round only the final displayed value, or to report the unrounded sum and let the reader see the small discrepancy. If you round each step and then add, the final cumulative value may be 0.99 or 1.01. If your table does not end at 1.00, check your rounding, not your data.

To calculate relative frequency, you need two numbers: the frequency of the value and the total number of observations. For example, if you have a dataset of 40 test scores and 8 of them are above 90, the relative frequency of scores above 90 is 8/40, which equals 0.20 or 20%. The calculation is always a division, never a subtraction or a percentage increase. A common error is to divide the total by the frequency, which gives a number larger than 1 and makes no sense. Another error is to use the frequency of a different category as the denominator. The denominator is always the total number of observations in the dataset, not the subtotal of a column. In a spreadsheet, you can compute this with a simple formula: if the frequency is in cell B2 and the total is in cell B10, then the relative frequency is =B2/$B$10. The dollar signs lock the denominator so it does not change when you copy the formula down.

Relative frequency vs probability is a distinction that confuses many students because the two are related but not the same. The probability of heads on any single flip is 0.50. The connection is the law of large numbers: as the number of trials increases, the relative frequency approaches the theoretical probability. But for a small sample, the relative frequency can be far from the probability. Some tables add a percent column, a cumulative frequency column, or a cumulative relative frequency column. The sum of the frequency column equals n, the total number of observations. If the table includes a cumulative relative frequency column, the last entry must be 1.00 (or 100%) because it includes all values. A well-built table labels every column clearly so the reader knows whether the numbers are counts or proportions.

One common error is the cumulative confusion: adding raw frequencies cumulatively and then dividing each cumulative sum by the total, which is correct, but forgetting to divide the last cumulative sum by itself, leaving a value other than 1.00. Another is the relative frequency mix-up: writing 25 instead of 0.25 in the relative frequency column, then wondering why the column does not sum to 1. A third is the empty bin error in spreadsheet functions like FREQUENCY in Excel or Google Sheets, where the user does not select enough cells for the array formula, causing a #N/A error. A fourth is the off-by-one bin edge: when binning continuous data, a value exactly equal to a bin boundary is counted in the wrong bin unless the bin is explicitly inclusive or exclusive. The fix for all these is to calculate from unrounded values and round only the final display. If your cumulative relative frequency column does not end at 1.00, the cause is almost always intermediate rounding, not a data error.

When the normal route is closed, check the unrounded relative frequencies first. If each relative frequency is 0.333333 and you wrote 0.33, the total will be 0.99, not 1.00. If your table software does not allow this, adjust the last cumulative entry to force the sum to 1.00 and add a footnote saying you rounded the last value. If the discrepancy is larger than a rounding error, such as a total of 0.85, then you have a counting error: either a frequency is wrong, the total is wrong, or a value is missing from the table. The relative frequency meaning depends on a correct denominator, and a single miscounted observation changes every row. If the table still will not reconcile, rebuild it from the raw data rather than patching the numbers.

What is relative frequency in simple terms?

Relative frequency is the count of a value divided by the total number of observations. If 5 out of 20 students are left-handed, the frequency is 5 and the relative frequency is 0.25.

Why do all relative frequencies sum to 1?

Because every observation is counted in exactly one category. Rounding intermediate values can make the total slightly off.

Can relative frequency be greater than 1?

No. A relative frequency of 1 means the value occurred in every observation. Values are always between 0 and 1 inclusive.

How many decimal places should I use for relative frequency?

Use one more decimal place than the raw data, or match the precision of the original measurements. Round only the final displayed value, not intermediate calculations, to avoid cumulative rounding errors.