Point Estimates and Margin of Error for a Mean

From a Sample Mean to an Estimate

In “Writing Full Responses About the Distribution of \(\bar{x}\),” you described how sample means vary from sample to sample. Now we use an observed sample mean for a different purpose: estimating an unknown population mean. The sample mean gives a single, useful estimate, but another random sample would usually produce a somewhat different value. Margin of error helps describe the precision of an estimate by expressing a distance around it.

A point estimate is a single statistic used to estimate a population parameter. For a population mean \(\mu\), the sample mean \(\bar{x}\) is the point estimate. As established in “Mean of the Sampling Distribution of x-bar,” the sample mean is an unbiased estimator of \(\mu\) under the sampling conditions described there. That does not mean every sample mean equals the population mean; it means the estimator’s sampling distribution is centered at \(\mu\).

Definition: The point estimate of a population mean \(\mu\) is the observed sample mean \(\bar{x}\). It is a single value calculated from sample data and used to estimate the population mean.

A point estimate is easy to report, but by itself it does not show how much estimates might vary across samples. For a quantitative variable, the sample standard deviation \(s\) describes the variability among the observations in the sample. To describe the estimated variability of the sample mean, we use the standard error, calculated as \(s/\sqrt{n}\). It estimates the standard deviation of the sampling distribution of \(\bar{x}\), which is \(\sigma/\sqrt{n}\) when the population standard deviation \(\sigma\) is known and the independence conditions allow that formula.

The standard error is not the margin of error. A margin of error multiplies the standard error by a critical multiplier that reflects the confidence procedure being used. In this tutorial, a multiplier will be supplied when needed. Later tutorials explain how to choose the appropriate multiplier for a one-sample \(t\) interval.

Formula: With a supplied critical multiplier \(c\), the margin of error for a mean is

\text{margin of error}=c\left(\frac{s}{\sqrt{n}}\right).

Here, \(s/\sqrt{n}\) is the estimated standard error of \(\bar{x}\), and \(c\) is the multiplier specified by the confidence procedure.

The margin of error is a distance in the same units as the original measurements. Starting at \(\bar{x}\), subtracting and adding that distance gives the endpoints of an interval:

\text{point estimate}\ \pm\ \text{margin of error} \qquad\text{or}\qquad \bar{x}\pm\text{margin of error}.

This interval gives a range of plausible values for the population mean under the procedure’s assumptions. A smaller margin of error means the interval is narrower and the estimate is more precise in this sense. It does not prove that the point estimate is close to \(\mu\), and it does not account for bias caused by a poor sampling method.

How the Pieces Fit Together

Identify the parameter and estimate.
Name the population mean \(\mu\) you want to estimate. The observed sample mean \(\bar{x}\) is the point estimate.

Find the estimated standard error.
Use the sample standard deviation and sample size: \(s/\sqrt{n}\). Keep \(s\) separate from the standard error; they describe different spreads.

Calculate the margin of error.
Multiply the estimated standard error by the supplied critical multiplier \(c\). Report the result in the measurement units.

Form and interpret the interval.
Place the margin of error on either side of \(\bar{x}\). Describe the interval as plausible values for the population mean, in context.

These calculations assume that the sample provides a reasonable basis for learning about the population. As in earlier tutorials on sampling, random selection and independence matter. If the sample is biased or unrepresentative, a small margin of error cannot repair that problem. Also, the shape and other conditions needed for a particular confidence procedure must be considered before using that procedure; the arithmetic alone does not establish that the interval is appropriate.

Worked Examples: Estimating a Population Mean

Worked Example: Time Spent Maintaining a Trail

A parks team takes a random sample of 25 weekly maintenance logs from a large set of logs. The sample mean time spent maintaining a trail is 6.4 hours, and the sample standard deviation is 1.5 hours. For an illustrative 95% confidence procedure, use the supplied multiplier \(c=2.064\). Find the point estimate, estimated standard error, margin of error, and interval for the population mean weekly maintenance time.

State. The parameter is \(\mu\), the population mean weekly trail-maintenance time. The point estimate is \(\bar{x}=6.4\) hours.

Plan. Use \(s/\sqrt{n}\) to estimate the standard error of the sample mean, then multiply by the supplied \(c\) to find the margin of error. The sample is described as random. The problem also describes a large set of logs, so the sample is not presented as a large fraction of a small finite population. We use the specified confidence procedure’s multiplier as given; selecting that multiplier is not part of this calculation.

Do. The estimated standard error is

\frac{s}{\sqrt{n}}=\frac{1.5}{\sqrt{25}} =\frac{1.5}{5}=0.30\text{ hours}.

The margin of error is

2.064(0.30)=0.6192\text{ hours}\approx0.62\text{ hours}.

The interval is the point estimate plus or minus the margin of error:

6.4\pm0.6192 \quad\Longrightarrow\quad (5.7808,\ 7.0192)\text{ hours},

or approximately \((5.78,\ 7.02)\) hours, with endpoints rounded to the nearest hundredth.

Conclude. The sample mean estimates the population mean weekly maintenance time as 6.4 hours. Using the stated procedure, the interval of about 5.78 to 7.02 hours gives plausible values for the population mean. The margin of error is about 0.62 hours, the distance from the point estimate to either endpoint.

Worked Example: Comparing Sample Sizes for Recharge Time

A technician samples rechargeable devices to estimate their mean recharge time. In one sample, \(n=36\), \(\bar{x}=82.5\) minutes, and \(s=12\) minutes. Use the supplied illustrative multiplier \(c=2\). Then compare the estimated standard error and margin of error with what they would be for a hypothetical sample of \(n=144\) if the sample standard deviation and multiplier stayed the same.

State. For the first sample, the point estimate of the population mean recharge time is 82.5 minutes.

Plan. Calculate \(s/\sqrt{n}\), then multiply by \(c\). For the comparison, keep \(s=12\) minutes and \(c=2\) fixed and change only \(n\). This isolates how sample size affects the estimated standard error and margin of error; it does not claim that a real second sample would have exactly the same sample standard deviation.

Do. For the sample of 36 devices,

\text{estimated standard error} =\frac{12}{\sqrt{36}} =\frac{12}{6}=2\text{ minutes},

so the margin of error is

2(2)=4\text{ minutes}.

The corresponding interval is \(82.5\pm4\), or \((78.5,\ 86.5)\) minutes.

For the hypothetical sample of 144 devices, still using \(s=12\) and \(c=2\),

\frac{12}{\sqrt{144}}=\frac{12}{12}=1\text{ minute}, \qquad \text{margin of error}=2(1)=2\text{ minutes}.

Conclude. With the sample standard deviation and multiplier held fixed, increasing the sample size from 36 to 144 reduces the estimated standard error from 2 minutes to 1 minute and the margin of error from 4 minutes to 2 minutes. The second margin of error is half as large because the sample size is four times as large. This comparison describes the formula’s behavior, not a guarantee that every larger sample produces a narrower interval under all circumstances.

Worked Example: A Point Estimate Is Not the Margin of Error

A school facilities team takes a random sample of 16 classrooms and records the number of minutes needed to reset each room between uses. The sample mean is 14.2 minutes, and the sample standard deviation is 4.8 minutes. For this example, use the supplied multiplier \(c=2.131\). Find the point estimate and margin of error, and explain what each number describes.

State. The population parameter is \(\mu\), the mean reset time for all classrooms in the population of interest. The point estimate is \(\bar{x}=14.2\) minutes.

Plan. First calculate the estimated standard error using \(s/\sqrt{n}\). Then multiply by \(c\) to find the margin of error. The sample is described as random, but a complete application of a confidence procedure would also require checking that procedure’s conditions. Here, the supplied multiplier lets us focus on what the estimate and margin of error mean.

Do. The estimated standard error is

\frac{s}{\sqrt{n}}=\frac{4.8}{\sqrt{16}} =\frac{4.8}{4}=1.2\text{ minutes}.

The margin of error is

2.131(1.2)=2.5572\text{ minutes}\approx2.56\text{ minutes}.

Using the unrounded margin of error, the interval is

14.2\pm2.5572 \quad\Longrightarrow\quad (11.6428,\ 16.7572)\text{ minutes},

or approximately \((11.64,\ 16.76)\) minutes.

Conclude. The sample mean of 14.2 minutes is the point estimate of the population mean reset time. The margin of error, about 2.56 minutes, is the distance used on either side of that estimate to form the interval. It is not the sample standard deviation, the estimated standard error, or a known measure of how far 14.2 is from the true population mean.

Common Mistakes and AP Exam Tips

Calling \(s\) the standard error. The sample standard deviation \(s\) describes spread among individual sample observations. The estimated standard error of \(\bar{x}\) is \(s/\sqrt{n}\). Show the division by \(\sqrt{n}\) before applying the multiplier.
Calling the point estimate the margin of error. The point estimate is \(\bar{x}\), a location. The margin of error is a distance. A complete response identifies them separately and gives units for both.
Adding the margin of error to only one side. The interval extends both below and above the point estimate: \(\bar{x}-\text{ME}\) to \(\bar{x}+\text{ME}\). Check that the point estimate is the midpoint of your endpoints.
Claiming that the margin of error guarantees accuracy. The margin of error describes uncertainty from sample-to-sample variation under a procedure. It does not measure bias or promise that the true mean is within a particular distance of the observed sample mean.
Forgetting context and units. A strong interpretation names the population mean and the measured quantity. For example, say “plausible values for the population mean weekly maintenance time,” not merely “the answer is 5.78 to 7.02.”
Treating a supplied multiplier as universal. Different confidence procedures can use different multipliers. If a problem supplies \(c\), use it as directed; do not assume that one value works for every sample size or confidence level.

A clear AP response makes the roles of the numbers visible: \(\bar{x}\) is the estimate, \(s/\sqrt{n}\) estimates the sampling variability of that estimate, and the multiplier turns that estimated standard error into a margin of error for the specified procedure. When you interpret the resulting interval, identify the population mean and keep the conclusion in the original units.

Key takeaway: The sample mean \(\bar{x}\) is a point estimate of the population mean \(\mu\). A margin of error is a distance calculated from a critical multiplier and the estimated standard error \(s/\sqrt{n}\); it extends on both sides of \(\bar{x}\) to form an interval.

Check Your Understanding

For each question, distinguish the point estimate, estimated standard error, and margin of error.

A random sample of 49 delivery routes has a mean completion time of 38 minutes and a sample standard deviation of 7 minutes. What is the point estimate of the population mean, and what is the estimated standard error?
A sample has \(n=25\), \(s=10\) liters, and a supplied multiplier \(c=2.1\). Calculate the estimated standard error and margin of error, including units.
A sample mean is 120 grams and the margin of error is 8 grams. Write the interval formed by the estimate and margin of error, and identify its midpoint.
Explain why a margin of error of 3 centimeters does not mean that the sample mean is known to be within 3 centimeters of the population mean.
If \(s\) and the supplied multiplier stay fixed while the sample size increases, what happens to \(s/\sqrt{n}\), and why does this affect the margin of error?