What Is Cumulative Frequency? How to Read and Calculate It

What Is Cumulative Frequency

You open a frequency distribution table and see the columns Frequency, Percent, Valid Percent, and Cumulative Percent. The last column is often the easiest to misread because its value increases down the rows, but it is not an independent percentage for each group. It is the cumulative percentage up to the group currently being examined.

Cumulative frequency is the running total of observations added from the first group through a specific group. When a variable's values are arranged in order, the cumulative frequency at value x tells you how many observations are less than or equal to x.

The basic formula is:

Fᵢ = f₁ + f₂ + ... + fᵢ

Here, fᵢ is the frequency of the i-th group, while Fᵢ is the cumulative frequency through the i-th group. To express it as a percentage, add the valid percentages of the groups cumulatively:

Pᵢ = Fᵢ / N × 100%

where N is the total number of valid observations. For example, a table has 20 people choosing level 1, 30 choosing level 2, and 50 choosing level 3. The cumulative frequency through level 2 is 20 + 30 = 50 people. The cumulative percentage through level 2 is 50 / 100 = 50%.

Cumulative frequency is meaningful only when the values are ordered. For a nominal variable such as field of study or province, adding values in the displayed order usually has no analytical meaning. It is more useful for ordinal variables, discrete quantitative variables, or income, age, and score intervals.

Why Cumulative Frequency Matters in Quantitative Research

In quantitative research, cumulative frequency gives you a quick view of the data distribution relative to a cutoff. You can answer questions such as how many people have a rating of 3 or below, what percentage of the sample has income no higher than 10 million VND, or how many students have used a service for a maximum of 2 years.

This measure also helps you identify the position of the median, quartiles, and percentiles. For example, if the total sample is 200 and you are locating the 50% position, the first row whose cumulative percentage reaches or exceeds 50% identifies the group containing the median. When you need more detail about the median, look at both the order of the groups and the cumulative frequency rather than relying on a single mean value.

Cumulative frequency also helps you check whether a data table is internally consistent. For a numerical cumulative frequency, the value must increase or remain unchanged from top to bottom, and the last row must equal the total number of valid observations. For Cumulative Percent, the last row is usually 100% when the table displays all valid groups.

In descriptive reporting, you can use cumulative frequency to describe the distribution pattern. For example, if 78% of respondents are no older than 30, that sentence is easier to understand than listing the frequency of every age group separately. Cumulative frequency does not, however, tell you about causes, causal relationships, or the magnitude of effects between variables.

You also need to distinguish it from measures of dispersion. Standard deviation shows how widely the data are dispersed around the mean, while variance is the squared form of deviation from the mean. These measures do not replace cumulative frequency.

What Cumulative Frequency Value Is Acceptable

Cumulative frequency does not have a general acceptable threshold like Cronbach's Alpha, KMO, or a p-value. It is a descriptive measure, so you evaluate it based on the total number of observations, the order of the groups, and the research question. A value of 60% is not automatically better than 40%. Its meaning depends on the cutoff you are examining.

You can use the following checks when reading the table:

The rows above are mathematical checks for a running total, not testing thresholds assigned by a research publication. You should therefore not write that cumulative frequency must be greater than 50% to be acceptable. The 50% point is used only to locate the median.

If you are checking data quality, also examine valid and missing values. A table with 250 data rows but only 230 valid responses will have Cumulative Percent based on 230, not 250. This differs from simply dividing the cumulative frequency by the total number of original rows.

How to Read Cumulative Frequency in the Output

In SPSS, create the table by going to Analyze > Descriptive Statistics > Frequencies. Move the variable you want to examine into Variable(s), keep Display frequency tables selected, and click OK. If the variable is numeric, check Value Labels and the order of the values before interpreting the output.

SPSS usually displays the following columns:

• Frequency: the number of observations belonging to the exact group or value in that row.
• Percent: the proportion of the group out of all rows, which may include missing values.
• Valid Percent: the proportion of the group out of valid observations, with missing values removed from the denominator.
• Cumulative Percent: the valid percentage accumulated through the current row.

Suppose you survey 100 people about their level of agreement on a 1 to 5 Likert scale and have no missing data. The shortened table below is illustrative output, not the result of an actual study:

In the row for level 3, Frequency is 29, meaning that 29 people selected exactly level 3. Cumulative Percent is 54.0, meaning that 54% of respondents selected level 1, 2, or 3. You should not read this value as 54% selecting level 3 alone.

In the row for level 4, Cumulative Percent is 85.0. This means that 85% of respondents selected a level from 1 through 4, or, equivalently, 15% selected level 5 in the illustrative table. If your research question concerns the group with agreement of 4 or above, you need to add groups 4 and 5, or recreate the table using an appropriate grouping scheme. Cumulative Percent is normally calculated from low to high, so it does not directly answer the question, “What percentage selected 4 or above?”

When missing values are present, you will see a Missing row in the Frequency section. Percent and Valid Percent can then differ. For example, with 90 valid responses and 10 missing values, a group with Frequency of 45 has Percent of 45%, but Valid Percent of 50%. Cumulative Percent uses Valid Percent, so read it together with Valid N.

For a continuous variable such as income or grade point average, SPSS may create too many rows if every value is different. Consider creating interpretable groups before running Frequencies. Dividing values into groups only to shorten the table can remove information, so state clearly how the intervals were created in the methods chapter.

What to Do When Cumulative Frequency Does Not Look Right

Calling cumulative frequency “unacceptable” is often imprecise. First identify whether you have a data error, an incorrect ordering, or a distribution that simply differs from your expectation. Work through the checks below in order.

Check the Value Order

If labels such as “Low,” “Medium,” and “High” were entered as text, SPSS may sort them alphabetically rather than according to their substantive order. Check Variable View, Value Labels, and the coding order. For an ordinal scale, codes 1, 2, and 3 are usually easier to control, but you must keep the coding consistent throughout the file.

Check Missing Values and Out-of-Range Codes

Some tables use 99 or 999 to represent “no response.” If this code has not been declared missing, SPSS will treat it as a real data group and distort the cumulative frequency. In Variable View, check the Missing column and compare it with the questionnaire coding guide.

Check Whether a Variable Needs Reverse Coding

For an item that requires reverse coding, the cumulative frequency of the individual item can still be calculated, but the trend will be interpreted incorrectly if you forget to reverse the coding. Cumulative frequency cannot correct a scale-design error. Review the codebook, create a new variable if necessary, and record the processing decision.

Regroup Values Only for a Methodological Reason

If a group contains very few observations, you may consider combining adjacent groups when the nature of the scale allows it. State which groups were combined and why. Do not delete rows with low frequencies simply to make a chart or table look more balanced.

Do Not Change Data to Reach a Desired Percentage

If the actual data show that most respondents fall into a low group, report that result. You can check validity, outliers, and missing values, but you must not change responses to make Cumulative Percent look better. When the data contain extreme values, review the guidance on outliers and record the processing criterion before removing an observation.

Distinguishing Cumulative Frequency, Percentage, and Frequency

Frequency is the number of times a value or group appears. In an SPSS table, it corresponds to Frequency. If 31 people select level 4, the frequency for level 4 is 31.

Percentage is commonly used for the proportion of a frequency out of the total number of observations. Depending on the source, it may be shown as a decimal or a percentage. In SPSS, Percent and Valid Percent are two ways to show the proportion, with different denominators when missing data are present.

Cumulative frequency is the total of the frequencies from the first group through the group being examined. Cumulative percentage is its proportional version. The three concepts can be compared as follows:

Cumulative frequency is different from the total frequency of an arbitrarily selected group. If you want to know the percentage from level 4 upward, do not mechanically use Cumulative Percent in the row for level 4. In a table ordered from low to high, you can calculate it as 100% minus the cumulative percentage through level 3, provided that missing values and the group order have been checked.

Common Errors

Reading Cumulative Percent as the Percentage of That Row Alone

This is the most common error. A Cumulative Percent of 70% in the row for level 4 does not mean that 70% of people selected level 4. Use Percent or Valid Percent to find the percentage for level 4 alone.

Forgetting to Check Missing Values

When Percent and Valid Percent differ, explain which denominator is being used. If you use Percent in your writing and ignore missing values, your conclusion about the respondent percentage may be wrong.

Accumulating a Nominal Variable

Accumulating gender, field of study, or province groups according to their displayed order usually has no statistical meaning. Cumulative frequency is more appropriate for a variable with a natural order.

Confusing Cumulative Frequency with a Cumulative Mean

Cumulative frequency counts the number or percentage of observations up to a cutoff. It is not a mean that is updated after each row. If you need to describe the center of the data, examine the mean, median, and appropriate measures of dispersion.

Reporting the Number Without Stating the Cutoff

The sentence “cumulative frequency is 54%” is not sufficiently clear. Write “54% of respondents have an agreement level from 1 to 3” so the reader knows how far the total has been accumulated.

Frequently asked questions

What is cumulative frequency in statistics?

Cumulative frequency is the total number of observations accumulated from the first group through a specific group. When the table is arranged in ascending order, it shows how many observations are less than or equal to the value being examined.

Is cumulative frequency the same as Cumulative Percent?

Cumulative Percent is cumulative percentage, meaning that cumulative frequency has been converted into a percentage of valid observations. Numerical cumulative frequency usually corresponds to the accumulated Frequency total, while Cumulative Percent is the percentage form.

What cumulative frequency value is acceptable?

This measure has no general acceptable threshold. Check whether it increases, whether the last row equals the total number of valid observations, and whether the final percentage is approximately 100%. The 25%, 50%, and 75% points are used to locate percentiles, not as data-quality standards.

Why are Percent and Valid Percent different in SPSS?

The two columns differ when the table contains missing data. Percent uses the total number of rows as the denominator, while Valid Percent removes missing observations from the denominator. Cumulative Percent is calculated from Valid Percent, so check the number of valid observations before writing the result.

Can cumulative frequency be used for gender?

SPSS can calculate it technically if gender is coded numerically, but a cumulative interpretation usually has no meaning because gender has no natural order. Use Frequency and Percent to describe each group rather than drawing a conclusion from the cumulative rows.

How should cumulative frequency be written in a thesis?

You can write: “The illustrative result shows that [X%] of respondents have [characteristic] no higher than [cutoff], corresponding to [F] valid observations. This percentage was identified from the Cumulative Percent column in the SPSS Frequencies table.” Replace the placeholders with the values in your file and state the cumulative cutoff clearly.

Open your SPSS file, run Analyze > Descriptive Statistics > Frequencies, and check the value order, missing values, and last row before placing the number in your results chapter. If you need to run statistics on your own .sav or .csv file and read the output step by step, see DoThesis M4 data analysis.