Population vs. Sample Mean: Demystifying the Notation

In our last post, we explored the three main ways to mea­sure cen­tral ten­den­cy: the mean, medi­an, and mode. We learned how to cal­cu­late these “aver­ages” to get a sin­gle num­ber that rep­re­sents a set of data.

How­ev­er, a crit­i­cal ques­tion soon aris­es: which data are we aver­ag­ing? Are we talk­ing about every sin­gle per­son in a group, or just a small piece of it?

This dis­tinc­tion is one of the most impor­tant con­cepts in sta­tis­tics. It’s the dif­fer­ence between a pop­u­la­tion and a sam­ple, and it’s why we have dif­fer­ent sym­bols (\mu vs. \bar{x}) for what looks like the same cal­cu­la­tion.

---

---

What is a Population?

In sta­tis­tics, a pop­u­la­tion is the entire, com­plete group of indi­vid­u­als, items, or data points you are inter­est­ed in study­ing. It’s the whole “enchi­la­da,” not just a slice.

• If you want to know the aver­age height of all men in Amer­i­ca, the pop­u­la­tion is all 150 mil­lion (or so) men in Amer­i­ca.
• If you want to know the aver­age lifes­pan of a spe­cif­ic brand of light­bulb, the pop­u­la­tion is every sin­gle light­bulb of that brand ever pro­duced.
• If you want to know the aver­age price of a house in New York City, the pop­u­la­tion is every sin­gle house in New York City.

The numer­i­cal val­ue that describes a char­ac­ter­is­tic of a pop­u­la­tion is called a para­me­ter. For exam­ple, the true aver­age height of all Amer­i­can men would be a pop­u­la­tion para­me­ter.

What is a Sample?

A sam­ple is a small, man­age­able sub­set tak­en from the pop­u­la­tion. Because it’s often impos­si­ble or imprac­ti­cal to study the entire pop­u­la­tion, we take a sam­ple and use it to con­clude the whole.

• Instead of mea­sur­ing all 150 mil­lion men, you might mea­sure 1,000 ran­dom­ly select­ed men. This is your sam­ple.
• Instead of test­ing every light­bulb until it burns out, you might test 500 bulbs. This is your sam­ple.
• Instead of apprais­ing every house, you might look at the sales data for 200 recent­ly sold hous­es. This is your sam­ple.

The numer­i­cal val­ue that describes a char­ac­ter­is­tic of a sam­ple is called a sta­tis­tic. The aver­age height of your 1,000-man sam­ple is a sam­ple sta­tis­tic.

In infer­en­tial sta­tis­tics, our main goal is to use a sam­ple sta­tis­tic (which we can cal­cu­late) to make an edu­cat­ed guess about a pop­u­la­tion para­me­ter (which we usu­al­ly don’t know).

Why Do We Need Samples? The Problem of Impracticality

Why don’t we just mea­sure the whole pop­u­la­tion? It would be more accu­rate, right?

Yes, but it’s almost always impos­si­ble or imprac­ti­cal.

Let’s go back to the “aver­age height of all men in Amer­i­ca” exam­ple. To find the true pop­u­la­tion mean, you would need to:

1. Find every sin­gle man in the coun­try at the exact same time.
2. Mea­sure all 150 mil­lion of them.
3. Add up all their heights and divide by 150 mil­lion.

This is not just dif­fi­cult; it’s phys­i­cal­ly impos­si­ble. By the time you fin­ished mea­sur­ing, new men would have been born, and oth­ers would have passed away. The pop­u­la­tion itself changes.

Even for a sta­t­ic pop­u­la­tion (like light­bulbs), it’s often imprac­ti­cal. To find the aver­age lifes­pan, you would have to test every bulb until it breaks, leav­ing you with no prod­ucts to sell.

This is why we sam­ple. Sam­ples are:

• Cost-effec­tive
• Time-effi­cient
• Prac­ti­cal

We accept a small amount of uncer­tain­ty in exchange for the abil­i­ty to get a good-enough answer.

Demystifying the Notation: Population Mean (\mu) vs. Sample Mean (\bar{x})

Because the dis­tinc­tion is so impor­tant, sta­tis­ti­cians use dif­fer­ent sym­bols to be clear about what they’re refer­ring to.

When we talk about the mean, this is the most com­mon con­fu­sion for begin­ners.

Con­cept	Pop­u­la­tion Mean	Sam­ple Mean
What is it?	The true aver­age of the entire pop­u­la­tion.	The aver­age of your sam­ple data.
Sym­bol	\mu (the Greek let­ter “mu”)	\bar{x} (read as “x‑bar”)
Type of Val­ue	A Para­me­ter: A fixed, unknown val­ue.	A Sta­tis­tic: A cal­cu­lat­ed, known val­ue.
Pur­pose	This is the val­ue we want to know or esti­mate.	This is the val­ue we use to esti­mate \mu.

The cal­cu­la­tion is the same (sum of val­ues / num­ber of val­ues), but the sym­bols tell you what data you’re using.

Decoding the Formulas: Understanding Sigma Notation

In sta­tis­tics books, you’ll see these for­mu­las writ­ten in a way that can look intim­i­dat­ing. Let’s break them down.

The fan­cy sym­bol \sum (the cap­i­tal Greek let­ter “Sig­ma”) is just a math­e­mat­i­cal instruc­tion that means “add every­thing up.”

Population Mean (\mu) Formula

\mu = \frac{\sum_{i=1}^{N} X_i}{N}

Let’s trans­late this:

• \mu: This is the pop­u­la­tion mean we want to find.
• \sum: “Add up…”
• X_i: “…each val­ue (X) in the pop­u­la­tion, start­ing from the first one (i=1) up to the last one (N).”
• N: “…and then divide by N, the total num­ber of items in the pop­u­la­tion.”

In plain Eng­lish: “Add up every sin­gle val­ue in the entire pop­u­la­tion and divide by the pop­u­la­tion size.”

Sample Mean (\bar{x}) Formula

\bar{x} = \frac{\sum_{i=1}^{n} x_i}{n}

Let’s trans­late this:

• \bar{x}: This is the sam­ple mean we are cal­cu­lat­ing.
• \sum: “Add up…”
• x_i: “…each val­ue (x) in our sam­ple, start­ing from the first one (i=1) up to the last one (n).”
• n: “…and then divide by n, the total num­ber of items in the sam­ple.”

In plain Eng­lish: “Add up all the val­ues in your sam­ple and divide by the sam­ple size.”

Key Con­ven­tion: N vs. n

You’ll notice that cap­i­tal N is used for the Pop­u­la­tion size (e.g., 150 mil­lion men), while low­er­case n is used for the Sam­ple size (e.g., the 1,000 men we mea­sured). This is a stan­dard con­ven­tion in sta­tis­tics.

The Importance of a Good Sample

The entire hope of this process is that our sam­ple mean (\bar{x}) is a good esti­mate of the pop­u­la­tion mean (\mu).

But what if our sam­ple is “skewed”?

If we want to find the aver­age height of Amer­i­can men, but we col­lect our sam­ple from a col­lege bas­ket­ball team, our \bar{x} is going to be very high. It will be a ter­ri­ble esti­mate of the true \mu. This is called sam­pling bias.

To get a good esti­mate, our sam­ple must be ran­dom and rep­re­sen­ta­tive of the pop­u­la­tion. This means every mem­ber of the pop­u­la­tion has an equal chance of being includ­ed in the sam­ple. (We’ll cov­er sam­pling tech­niques in a future post!)

---

Key Takeaways

• A Pop­u­la­tion (N) is the entire group you want to study.
• A Sam­ple (n) is a small sub­set of that group.
• We use sam­ples because study­ing the entire pop­u­la­tion is usu­al­ly impos­si­ble or imprac­ti­cal.
• The Pop­u­la­tion Mean (\mu) is a fixed, unknown para­me­ter we want to find.
• The Sam­ple Mean (\bar{x}) is a cal­cu­lat­ed sta­tis­tic we use to esti­mate \mu.
• The for­mu­las are the same (sum divid­ed by count), but the nota­tion (\mu vs. \bar{x} and N vs. n) is dif­fer­ent to show this crit­i­cal dis­tinc­tion.

---

Frequently Asked Questions (FAQ)

1. Why don’t we just call both of them the “mean”?

We do! But we must spec­i­fy which mean. \mu is the pop­u­la­tion mean, and \bar{x} is the sam­ple mean. Being pre­cise is crit­i­cal in sta­tis­tics to avoid con­fu­sion.

2. Is the sam­ple mean (\bar{x}) always close to the pop­u­la­tion mean (\mu)?

Not always, espe­cial­ly with small or biased sam­ples. The goal of good sta­tis­ti­cal prac­tice is to col­lect a sam­ple that is large enough and ran­dom enough to ensure \bar{x} is a reli­able esti­mate of \mu.

3. What does x_i mean in the for­mu­la?

The ‘i’ is an “index” vari­able. It’s just a place­hold­er. x_i means “the i-th val­ue in the dataset.” So x_1 is the first val­ue, x_2 is the sec­ond, and so on, up to x_n (the last val­ue in the sam­ple).