Back to homeMathematics · Integrated Math I · Change subject
Start for free
Math I · Mathematics
Statistics · Statistics

Populations, Samples, and Sampling Methods

In 1936, a magazine called The Literary Digest asked more than 2 million Americans who they would vote for in the presidential election. It confidently predicted that Alf Landon would win in a landslide. Franklin Roosevelt won in a landslide instead. How can a poll with millions of answers be so wrong? Because the size of a sample is not everything: how you choose who to ask matters more than how many you ask.

Population, sample, and census

Statistics starts with a question about a group. You rarely have time or money to ask everyone, so you look at a smaller piece of the group and use it to learn about the whole.

A population of N individuals drawn as dots in a grid, with a sample of size n = 4 highlighted. Population: N individuals sample size n = 4
Practice★☆☆Not started
An electronic components factory produces a batch of $5000$ resistors. To check their quality, the control department selects and analyzes $150$ resistors. Identify the…

An electronic components factory produces a batch of $5000$ resistors. To check their quality, the control department selects and analyzes $150$ resistors. Identify the population, the sample, the individual and the sample size in this statistical study.

Solve
Practice★☆☆Not started
In a municipality of $24000$ inhabitants, a study on reading habits is to be carried out. To do this, $800$ citizens are interviewed. Determine what the study population…

In a municipality of $24000$ inhabitants, a study on reading habits is to be carried out. To do this, $800$ citizens are interviewed. Determine what the study population is, what the sample is and what the sample size is.

Solve
Practice★☆☆Not started
A software company wants to know its users' opinions on a new update of its application. Briefly explain in what situation the company would carry out a census and in…

A software company wants to know its users' opinions on a new update of its application. Briefly explain in what situation the company would carry out a census and in what situation it would carry out a sample study.

Solve

Sample size and percent

The sample is a fraction of the population, and we often describe that fraction as a percent. This is just the percent equation you already know, with "part" = $n$ and "whole" = $N$.

A long bar for a population of 2,500 trout with a small piece representing the 8 percent sample. Population: N = 2,500 trout Sample: 8% of 2,500, so n = 200
Practice★☆☆Not started
$3400$ deer live in a nature reserve. A team of biologists decides to take a sample equivalent to $15\%$ of the population to carry out a health study. Calculate the…

$3400$ deer live in a nature reserve. A team of biologists decides to take a sample equivalent to $15\%$ of the population to carry out a health study. Calculate the sample size they must select.

Solve
Practice★☆☆Not started
To carry out a quality control in a soft drinks factory, $45$ cans are extracted from a total production of $1800$ cans per day. What percentage of the total production…

To carry out a quality control in a soft drinks factory, $45$ cans are extracted from a total production of $1800$ cans per day. What percentage of the total production does the selected sample represent?

Solve
Practice★★☆Not started
A polling company has selected a sample of $630$ people for a market study. If it is known that this sample represents exactly $3,5\%$ of the target population,…

A polling company has selected a sample of $630$ people for a market study. If it is known that this sample represents exactly $3,5\%$ of the target population, calculate the total size of the study population.

Solve

Types of variables

A variable is whatever you record about each individual. Before you can summarize data, you need to know what kind of variable you have.

Tree diagram sorting variables into qualitative (nominal or ordinal) and quantitative (discrete or continuous). Variable Qualitative (categorical) Quantitative (numerical) nominal ordinal discrete continuous
Type Values Examples
Qualitative, nominal Categories with no order Phone brand, fur color, favorite pizza topping
Qualitative, ordinal Categories with a natural order Risk level (low/medium/high), T-shirt size (S/M/L), letter grade
Quantitative, discrete Numbers you count (gaps between possible values) Siblings, moons of a planet, heartbeats in a minute
Quantitative, continuous Numbers you measure (any value in an interval) Weight, height, time, temperature, volume
Practice★☆☆Not started
Classify the following statistical variables into qualitative, discrete quantitative or continuous quantitative: The fur color of a fox species. The number of natural…

Classify the following statistical variables into qualitative, discrete quantitative or continuous quantitative:

  • The fur color of a fox species.

  • The number of natural satellites of the planets in the solar system.

  • The mass in pounds of postal shipments from a parcel company.

  • The brand of the computers in a computer lab.

Solve
Practice★☆☆Not started
State whether the following quantitative variables are discrete or continuous: a) The volume of water in a reservoir measured in cubic feet. b) The number of heartbeats…

State whether the following quantitative variables are discrete or continuous:

  1. a) The volume of water in a reservoir measured in cubic feet.
  2. b) The number of heartbeats per minute of an athlete after a race.
  3. c) The maximum daily temperature recorded in a city.
  4. d) The number of passengers on a commercial flight.
Solve
Practice★☆☆Not started
A team of nutritionists measures the amount of sugar, in grams, contained in different brands of cereals from a supermarket. Identify what the statistical variable of…

A team of nutritionists measures the amount of sugar, in grams, contained in different brands of cereals from a supermarket. Identify what the statistical variable of the study is and what type it is.

Solve

Variable type and the "center" of the data

The type of variable tells you which summary makes sense.

  • Nominal: you can only report the mode (the most common category). There is no "average phone brand."

  • Ordinal: the median (middle category) or the mode make sense.

  • Quantitative: the mean and the median both make sense, but they behave differently when there are outliers (extreme values).

Try changing the largest value in the widget and watch which summary moves.

Practice★★☆Not started
In a marine ecology study, the following characteristics of a dolphin population are recorded: a) Species to which it belongs. b) Body length in feet. c) Number of…

In a marine ecology study, the following characteristics of a dolphin population are recorded:

  1. a) Species to which it belongs.
  2. b) Body length in feet.
  3. c) Number of offspring it has had throughout its life.
  4. d) Extinction risk level (Low, Vulnerable, Endangered).

Classify each statistical variable in detail. In the case of qualitative variables, state explicitly whether they are ordinal or nominal.

Solve
Practice★★★Not started
In an astronomical observatory, the following characteristics of an exoplanet catalog are recorded: A) Distance to the host star (in Astronomical Units). B) Planet type…

In an astronomical observatory, the following characteristics of an exoplanet catalog are recorded:

A) Distance to the host star (in Astronomical Units).

B) Planet type (Rocky, Gas giant, Ice giant).

C) Number of confirmed moons.

D) Estimated mean temperature (in Kelvin).

Classify each of these variables rigorously. For the quantitative variables, determine in which cases it makes statistical sense to calculate the arithmetic mean and in which the median could be a more representative measure of central tendency if extreme outliers existed.

Solve

Random and stratified sampling

A good sample looks like the population in miniature. The safest way to get one is to let chance decide.

Sometimes the population splits into groups (strata) that may behave differently: grade levels in a school, departments in a company. A simple random sample could, by bad luck, include very few people from one group. Stratified sampling fixes that: you take a random sample inside each stratum, with a size proportional to the stratum.

Stratified sampling: a population of 600 students split into three grades, and a proportional sample of 60 students. Population (N = 600) Grade 6: 240 Grade 7: 200 Grade 8: 160 Sample (n = 60) n₁ = 24 n₂ = 20 n₃ = 16

You can explore proportional splits with this widget. Each number in partes is a stratum size.

Practice★★☆Not started
In a high school there are $120$ students in 9th grade, $100$ in 10th grade, $90$ in 11th grade, and $50$ in 12th grade. A sample of $72$ students is to be selected…

In a high school there are $120$ students in 9th grade, $100$ in 10th grade, $90$ in 11th grade, and $50$ in 12th grade. A sample of $72$ students is to be selected using stratified random sampling with proportional allocation. Calculate how many 11th-grade students must be part of the sample.

Solve
Practice★★☆Not started
A poultry farm has three sheds with laying hens: shed A has $4500$ hens, shed B has $3000$ and shed C has $2500$. A sample of $200$ hens is to be extracted using…

A poultry farm has three sheds with laying hens: shed A has $4500$ hens, shed B has $3000$ and shed C has $2500$. A sample of $200$ hens is to be extracted using stratified sampling with proportional allocation. Determine the sample size that corresponds to each of the three sheds.

Solve
Practice★★☆Not started
The following diagram represents a stratified sampling process with proportional allocation. If the total population consists of $2400$ individuals, calculate the…

The following diagram represents a stratified sampling process with proportional allocation.

POPULATION 40% 25% ? SAMPLE n₁ = ? n₂ = ? n₃ = ?

If the total population consists of $2400$ individuals, calculate the percentage and the population size of the orange stratum. If the total sample is $160$ individuals, determine how many individuals of each color make up the sample.

Solve

Working backward from a stratified sample

Because every stratum is sampled at the same rate, you can set up a proportion and solve for whatever is missing:

$$\frac{n_i}{n} = \frac{N_i}{N}$$

The share of the sample equals the share of the population.

Practice★★★Not started
In a municipality divided into three districts (North, South and Center), a survey has been carried out using stratified sampling with proportional allocation. The total…

In a municipality divided into three districts (North, South and Center), a survey has been carried out using stratified sampling with proportional allocation. The total sample is $300$ people, of which $120$ belong to the North district and $75$ to the South district. If the total population of the municipality is $18000$ inhabitants, calculate the population of each of the three districts.

Solve
Practice★★★Not started
In a technology company, the workforce is divided into three categories: $50\%$ are programmers, $30\%$ are designers and the rest are managers. A work climate study is…

In a technology company, the workforce is divided into three categories: $50\%$ are programmers, $30\%$ are designers and the rest are managers. A work climate study is carried out by selecting a sample using stratified sampling with proportional allocation. If there are exactly $16$ managers in the sample, what is the total sample size and how many programmers and designers make it up?

Solve
Practice★★★Not started
A factory has two machines, A and B, for the production of gears. It is known that machine A produces $65\%$ of the daily total. The quality department takes a…

A factory has two machines, A and B, for the production of gears. It is known that machine A produces $65\%$ of the daily total. The quality department takes a stratified sample with proportional allocation of $180$ gears to evaluate the cutting precision. If machine B produces $1400$ gears a day, calculate the total daily population size and the number of gears from each machine that make up the sample.

Solve

Stratified sampling with unknowns

Sometimes the stratum sizes are algebraic expressions. Then the proportion becomes an equation in $x$, which you solve with the linear-equation skills you already have.

A bar split into three strata labeled with expressions in x: 3x, 2x plus 20, and x plus 40. The whole bar is 6x plus 60. Chess: 3x Robotics: 2x + 20 Art: x + 40 Population: 6x + 60
Practice★★★Not started
In a study on the adoption of a new technology, the population of a company is divided into three departments, whose workforce sizes are given by algebraic expressions…

In a study on the adoption of a new technology, the population of a company is divided into three departments, whose workforce sizes are given by algebraic expressions in terms of $x$. A sample is taken using stratified random sampling with proportional allocation of size $130$.

Sales N₁ = 3x Marketing N₂ = x + 40 Development N₃ = 2x - 20 SAMPLE n = 130 n₁ = 50 n₂ = ? n₃ = ?

Knowing that there are exactly $50$ employees from the Sales department in the sample, set up an equation to determine the value of $x$, calculate the total population of the company and the sample size corresponding to the Marketing and Development departments.

Solve

Systematic sampling

Another unbiased way to pick a sample: number the population from $1$ to $N$ and select every $k$th individual, starting from a randomly chosen position.

A list of 20 individuals in a row. Positions 3, 8, 13, and 18 are highlighted, each 5 apart. N = 20, n = 4, so k = 20 ÷ 4 = 5 k = 5 a = 3a + k = 8a + 2k = 13a + 3k = 18
Practice★★☆Not started
On an assembly line, $1200$ cell phones have been manufactured in one day. A quality control is to be carried out by selecting a sample of $40$ phones using systematic…

On an assembly line, $1200$ cell phones have been manufactured in one day. A quality control is to be carried out by selecting a sample of $40$ phones using systematic sampling. If the first phone selected at random is the one occupying position $14$, determine the sampling interval, and the positions of the first three and the last phone selected in the sample.

Solve
Practice★★★Not started
A database contains the records of $N = 2520$ clients. It is desired to extract a sample of size $n = 45$ using systematic sampling. The random start determines that the…

A database contains the records of $N = 2520$ clients. It is desired to extract a sample of size $n = 45$ using systematic sampling. The random start determines that the first client selected is the one occupying position $a = 23$.

Find a general algebraic expression for the position of the $i$-th individual in the sample (for $1 \le i \le 45$) and determine what position in the database is occupied by the client who was selected in the $32$-th place in the sample.

Solve

Cluster sampling vs. stratified sampling

Stratified sampling is not the only way to use groups. In cluster sampling, you split the population into groups called clusters, randomly choose a few whole clusters, and survey everyone in the chosen ones.

Two panels with four groups each. Cluster sampling surveys everyone in two randomly chosen groups. Stratified sampling surveys a few individuals in every group. Cluster sampling Stratified sampling all surveyed not chosen not chosen all surveyed
Practice★★★Not started
The educational authorities of a region want to assess the English level of 10th-grade students. In the region there are $25$ high schools, each with exactly $120$…

The educational authorities of a region want to assess the English level of 10th-grade students. In the region there are $25$ high schools, each with exactly $120$ students enrolled in that year. A total sample of $600$ students is required. Two sampling methods are proposed:

  • Method A: Choose $5$ high schools at random and assess all their 10th-grade students.

  • Method B: Choose $24$ 10th-grade students at random from each of the $25$ high schools.

Identify the type of sampling applied in each method and argue which of the two designs guarantees greater representativeness of the total population, explaining the possible disadvantages of the method you discard.

Solve

Biased samples

A sample is representative if it mirrors the population. A sample is biased if the way it was chosen systematically favors some part of the population. A biased sample gives wrong answers no matter how large it is.

A biased selection method produces millions of replies that are still not representative of the population. Biased selection method Millions of repliesfrom the wrong people Size cannot fix a biased method
Practice★★☆Not started
A city council wants to know citizens' opinions on the construction of a new bike lane downtown. To do this, a pollster stands outside a bicycle shop and interviews…

A city council wants to know citizens' opinions on the construction of a new bike lane downtown. To do this, a pollster stands outside a bicycle shop and interviews $150$ people leaving it. Analyze whether the selected sample is representative of the city's population and explain what type of bias is being committed in the study design.

Solve
Practice★★☆Not started
It is desired to estimate the mean daily time spent on social media by teenagers across an entire state. To do this, $200$ students are randomly selected from a private…

It is desired to estimate the mean daily time spent on social media by teenagers across an entire state. To do this, $200$ students are randomly selected from a private high school located in the most affluent neighborhood of the capital city. Justify why this sample is not representative of the population and propose a more appropriate sampling method.

Solve
Practice★★★Not started
A national television channel wants to predict the outcome of an important legislative referendum. To do this, during a prime-time debate program, it asks viewers to…

A national television channel wants to predict the outcome of an important legislative referendum. To do this, during a prime-time debate program, it asks viewers to call a premium-rate phone number or send an SMS to vote for their preferred option. More than $100000$ responses are received.

Identify and explain at least three sources of bias (systematic errors) in this sampling design and reason why, despite the huge sample size, the results are probably not representative of the general population.

Solve

Back to the start: The Literary Digest built its list from telephone directories and car registrations, which in 1936 left out many poorer voters (undercoverage), and only about a quarter of the people who got a ballot mailed it back (voluntary response). Meanwhile, pollsters who used much smaller, carefully chosen samples predicted Roosevelt correctly. More than 2 million replies could not fix a biased method.

Summary

See all the exercises in this section