Statistics Frequency Distributions Introductory Statistics 10 min read April 25, 2026
BY: Statistics Fundamentals Team
Reviewed By: Minsa A (Senior Statistics Editor)

Computed Mean vs Actual Mean: The Complete Statistics Guide

You calculated the mean from a frequency table. Now a problem hands you an "actual mean" and asks you to compare the two. What does that comparison tell you, and how do you decide which answer choice is correct?

This guide breaks down what each type of mean is, exactly how to calculate the computed mean step by step, and how to use the 5% rule to determine whether your result is close to the actual mean — with three fully worked examples.

What Is the Computed Mean and What Is the Actual Mean?

These two terms show up together in textbook problems where data has been summarized into a frequency distribution rather than listed out individually. They measure the same thing — the average — but they come from different sources, and that source gap is exactly what the comparison is testing.

The computed mean (also called the grouped mean or estimated mean) is calculated using the class intervals and frequencies in the table. Because individual values inside each interval are unknown, you substitute the midpoint of each class and treat it as representative of every value in that group.

The actual mean is the true arithmetic average calculated from every individual data value in the original, ungrouped dataset. In textbook problems, it is given to you so you can check how accurate your computed result turned out to be.

💡
Quick Definition

Computed mean = average estimated from grouped data using midpoints. Actual mean = precise average from all individual data points. The difference between them reflects how much information is lost when data is summarized into intervals.

🔑 Key Takeaways

The essential ideas you need before working through any computed-vs-actual-mean problem.

The computed mean uses midpoints as approximations. This is why it differs from the actual mean — midpoints rarely match the true center of each class's values.

The 5% threshold decides "close" vs "not close." If the difference is less than 5% of the actual mean, the means are considered close.

Only two of the four answer choices can ever be correct. Options pairing "close" with "> 5%" or "not close" with "< 5%" are logically contradictory and always wrong.

Larger samples reduce the gap. As sample size grows, the computed mean converges toward the actual mean — a direct consequence of the Law of Large Numbers.

The formula never changes. Σ(f × x) ÷ Σf works for every grouped dataset, regardless of topic — temperatures, speeds, test scores, or anything else.

What Is the Computed Mean?

When raw data is organized into a frequency distribution, the individual values inside each class interval are hidden. A temperature reading in the 45–49°F interval could be 45.1, 47.3, or 48.9 — the table does not say. To calculate any kind of average from this grouped format, you need a single representative value for each class.

That representative is the midpoint: the number exactly halfway between the lower and upper class limits. Using midpoints to estimate the mean is a standard technique in descriptive statistics, and the resulting value is the computed mean.

Computed Mean = Σ(f × x) / Σf
where x = class midpoint, f = class frequency

The computed mean gives you a workable estimate when only summarized data exists. It's used by meteorologists analyzing temperature range tables, traffic engineers working with speed distribution data, and census analysts working with income brackets — any situation where grouped summaries stand in for raw records.

How to Find the Midpoint of a Class

The midpoint formula is always the same:

Midpoint = (Lower Class Limit + Upper Class Limit) / 2

For a class of 40–44, the midpoint is (40 + 44) / 2 = 42. For 45–49, it is (45 + 49) / 2 = 47. This pattern holds regardless of interval width.

What Is the Actual Mean?

The actual mean is the straight arithmetic average of every individual value in the dataset. It uses no shortcuts or approximations.

Population Mean: μ = Σx / N
Sample Mean: x̄ = Σx / n

In textbook problems, the actual mean is provided as a benchmark. The exercise is asking: given that you only had the frequency table, how close did your estimate get to the truth? This is genuinely useful to know. If the computed mean regularly falls within a small margin of the actual mean, frequency distributions are a reliable summary tool for that type of data.

Term Symbol Source When Used
Computed Mean c Frequency table using midpoints When only grouped data is available
Actual Mean x̄ or μ Every individual data value Given in problems as the reference point
Sample Mean Random sample of raw values Most research and surveys
Population Mean μ Every member of the population Census data, complete records

How to Calculate the Computed Mean From a Frequency Distribution

Six steps. They apply to every grouped-data problem — temperature, speed, test scores, anything.

1
List the class intervals and their frequencies

Read these directly from the frequency distribution table given in the problem.

2
Find the midpoint of each class

Use: Midpoint = (Lower Limit + Upper Limit) / 2

3
Multiply each midpoint by its frequency (f × x)

Do this for every row in the table separately.

4
Sum all the f × x products: Σ(f × x)

Add the results from every row into a single total.

5
Sum all frequencies: Σf

Add up all frequency values. This is your total number of data points.

6
Divide: Computed Mean = Σ(f × x) / Σf. Round to the nearest tenth.

This is your final answer for part one of the problem.

Worked Example

Temperature Data — Frequency Distribution

Given frequency distribution of low temperatures (°F):

Low Temp (°F) Frequency (f) Midpoint (x) f × x
40–44 3 42 126
45–49 4 47 188
50–54 9 52 468
55–59 5 57 285
60–64 2 62 124
Total 23 1,191

Computed Mean = 1,191 / 23 = 51.8°F

If the problem tells you the actual mean is 51.8°F, you already know the result before moving to the comparison step.

How to Compare the Computed Mean to the Actual Mean — The 5% Rule

This is the section that determines which answer choice you select. The rule used in Triola's Elementary Statistics (the most widely assigned introductory statistics textbook) treats the computed mean as "close" to the actual mean when their difference is less than 5% of the actual mean.

% Difference = |Computed Mean − Actual Mean| / Actual Mean × 100

Run this calculation, then apply the decision:

The 5% Decision Rule
Result
Verdict
Reason
% Difference < 5%
Computed mean IS close
Difference is less than 5% of actual mean
% Difference ≥ 5%
Computed mean is NOT close
Difference is more than 5% of actual mean

The Four Answer Choices Decoded

Every version of this multiple-choice question uses the same four options. Two can never be correct. Here is what each one means and when to pick it:

✓ CAN BE CORRECT
Close — difference < 5%

Pick this when % difference is under 5%. The grouped data gave a good estimate.

✓ CAN BE CORRECT
Not close — difference > 5%

Pick this when % difference is 5% or higher. The midpoint approximation introduced too much error.

✗ ALWAYS WRONG
Close — difference > 5%

Self-contradictory. A difference over 5% cannot make the means "close" by definition.

✗ ALWAYS WRONG
Not close — difference < 5%

Self-contradictory. A difference under 5% cannot make the means "not close" by definition.

⚠️
Common Mistake

Students sometimes compute 5% of the computed mean instead of 5% of the actual mean. The rule always uses the actual mean as the reference. Double-check which number you divide by.

Three Practice Problems With Full Solutions

Practice Problem 1 — Close Result

Speed Data: Compare to Actual Mean of 46.6 mph

Speed (mph) Frequency (f) Midpoint (x) f × x
42–452643.51,131
46–491347.5617.5
50–53751.5360.5
54–57355.5166.5
58–61159.559.5
Total502,335

Step 1 — Computed Mean: 2,335 / 50 = 46.7 mph

Step 2 — Difference: |46.7 − 46.6| = 0.1

Step 3 — 5% of actual mean: 0.05 × 46.6 = 2.33

Step 4 — Compare: 0.1 < 2.33

% difference = 0.1 / 46.6 × 100 = 0.21%

Answer: D — The computed mean is close to the actual mean because the difference is less than 5%

The grouped speed data produced an estimate just 0.21% away from the true mean. That is a tight result.

Practice Problem 2 — Not Close Result

Temperature Data: Compare to Actual Mean of 57.2°F

Low Temp (°F) Frequency (f) Midpoint (x) f × x
40–4414242
45–4914747
50–54452208
55–59457228
60–64962558
Total191,083

Step 1 — Computed Mean: 1,083 / 19 = 57.0°F

Step 2 — Difference: |57.0 − 57.2| = 0.2

Step 3 — 5% of actual mean: 0.05 × 57.2 = 2.86

Step 4 — Compare: 0.2 < 2.86

% difference = 0.2 / 57.2 × 100 = 0.35%

Answer: B — The computed mean is close to the actual mean because the difference is less than 5%

Despite the data being skewed toward the higher temperature ranges, the midpoint method still produced a close estimate.

Practice Problem 3 — True Not-Close Result

Speed Data: Compare to Actual Mean of 50.9 mph

Speed (mph) Frequency (f) Midpoint (x) f × x
42–452943.51,261.5
46–491347.5617.5
50–53651.5309
54–57455.5222
58–61159.559.5
Total532,469.5

Step 1 — Computed Mean: 2,469.5 / 53 = 46.6 mph

Step 2 — Difference: |46.6 − 50.9| = 4.3

Step 3 — 5% of actual mean: 0.05 × 50.9 = 2.545

Step 4 — Compare: 4.3 > 2.545

% difference = 4.3 / 50.9 × 100 = 8.45%

⚠️
Answer: D — The computed mean is NOT close to the actual mean because the difference is more than 5%

The heavy skew toward lower speeds (29 out of 53 readings in the 42–45 range) pulled the midpoint estimate well below the true mean. The grouped data is a poor approximation here.

Why the Computed Mean Differs From the Actual Mean

The core reason is that midpoints are substitutions, not facts. When you assign the value 47 to every observation in the 45–49 interval, you are assuming each reading falls exactly at the center of that range. Real data rarely behaves that neatly.

If most of the values in a 45–49 temperature class actually cluster around 48 or 49, the midpoint of 47 underestimates the average for that group. Multiply that small error across multiple classes and it accumulates into a meaningful gap between computed and actual means.

Three factors drive how large the gap gets:

  • Distribution shape — symmetric distributions produce smaller errors because values tend to balance around the midpoint. Skewed distributions accumulate error in the tail direction.
  • Class interval width — wider intervals give midpoints less precision. Narrower intervals produce better estimates.
  • Sample size — with more observations, midpoint errors tend to cancel out. This is why larger datasets produce computed means closer to the actual mean.
< 2%
Typical midpoint error for symmetric distributions
5%+
Error threshold for "not close" determination
∞ → μ
Computed mean converges to actual mean as n grows

Where Grouped Data Means Show Up in Real Life

The computed-vs-actual comparison is not just a textbook exercise. Every field that reports data in ranges uses some version of this calculation, and the accuracy of the estimate has real consequences.

The U.S. Census Bureau publishes household income in $5,000 brackets. Analysts who need average income for a region must compute it from those brackets using midpoints — exactly the same method. The actual mean, if recoverable from raw tax records, serves as the accuracy check.

Traffic safety engineers work with speed distributions sorted into 5 mph intervals from radar guns. The computed mean of that distribution feeds directly into speed limit decisions and road design. An 8% error versus the actual mean could meaningfully change the recommendation.

Meteorologists report historical temperatures in ranges, not individual readings. Climate trend analysis relies on computing means from these grouped records, then validating against station averages (the actual mean) when those are available.

In each case, the 5% check is a practical quality test: is this summary close enough to trust, or does the grouping introduce too much distortion?

Self-Check Before Submitting Your Answer

  • Did you find the midpoint of every class, not just use the lower limit?
  • Did you multiply midpoint × frequency for each row separately, before summing?
  • Did you divide by total frequency, not by the number of classes?
  • Did you calculate % difference using the actual mean as the denominator, not the computed mean?
  • Did you round the computed mean to the nearest tenth as directed?
  • Did you check whether the answer options pair "close" with "< 5%" and "not close" with "> 5%"?

FAQs

The computed mean is estimated from grouped data using class midpoints. The actual mean is the precise arithmetic average calculated from every individual data value. The computed mean is an approximation; the actual mean is the truth.

When the absolute difference between them is less than 5% of the actual mean. Calculate % difference = |computed − actual| / actual × 100. A result under 5% means the means are close.

Because the individual values inside each class interval are unknown in grouped data. The midpoint is the single best estimate for the average of all values within that range, so it substitutes for the missing raw readings.

Any option that says "close because the difference is more than 5%" or "not close because the difference is less than 5%" is always logically wrong. Only "close because < 5%" and "not close because > 5%" can be correct.

Theoretically yes — if every value in each class interval happened to fall exactly at its midpoint. In practice this is rare, but for symmetric distributions with many observations, the difference can be extremely small.

Yes. Larger samples allow midpoint approximation errors to average out across more observations. This is a practical expression of the Law of Large Numbers, which guarantees that the sample mean converges toward the population mean as n increases.

Always 5% of the actual mean. The actual mean is the reference point — the thing you are checking your estimate against. Using the computed mean as the denominator would change the threshold and give you a wrong answer.