Understanding Populations, Random Sampling, and Bias - Seventh Grade Mathematics

Statistical investigation relies on representative random sampling to draw valid inferences about large populations without the prohibitive cost and logistical challenge of a full census.

Populations versus Samples

In statistics, a population refers to the complete, entire collection of individuals, objects, or measurements that you wish to investigate. A sample is a smaller subset selected from that population to gather empirical data.

               POPULATION vs SAMPLE
   +-------------------------------------------------------+
   | POPULATION: The entire group of interest              |
   | (e.g., All 1,200 students in a middle school)         |
   |                                                       |
   |        +-----------------------------------+          |
   |        | SAMPLE: The subset chosen to test |          |
   |        | (e.g., 60 randomly chosen students|          |
   |        +-----------------------------------+          |
   +-------------------------------------------------------+

Surveying every single member of a population is called conducting a census. While a census provides complete information, it is often prohibitively expensive, time-consuming, or physically impossible. A well-designed sample allows statisticians to make highly accurate predictions about the entire population using only a fraction of the time and resources.

Representative versus Biased Sampling

For a sample to produce valid, trustworthy conclusions about a population, it must be representative. A representative sample accurately reflects the diverse characteristics, demographics, and proportions of the overarching population.

               REPRESENTATIVE vs BIASED SAMPLING
   Representative Sample:  Mirrors the population; results are VALID.
   Biased Sample:          Favors certain outcomes or groups; results are INVALID.

A sample is biased if some members of the population are systematically more likely to be selected than others, or if the method of collection distorts the true opinions or characteristics of the group.

Common Types of Sampling Bias

Biased sampling methods lead to flawed conclusions that do not generalize to the population:

+-----------------------+-----------------------------+------------------------+
| Sampling Method       | How It Works                | Why It Is Biased       |
+-----------------------+-----------------------------+------------------------+
| Convenience Sample    | Selects individuals who are | Excludes people who are|
|                       | easiest to reach or find    | not in that location   |
| Voluntary Response    | Participants choose whether | Attracts individuals   |
|                       | to participate (call-in/web)| with extreme opinions  |
+-----------------------+-----------------------------+------------------------+

For example, asking the first twenty students entering the school band room what their favorite school elective is represents a convenience sample. It will heavily over-represent band students and fail to reflect the general student body.

Random Sampling Techniques

The gold standard for achieving a representative sample is random sampling. In a random sample, every individual member of the population possesses an equal, non-zero probability of being selected.

               RANDOM SAMPLING METHODS
   1. Simple Random Sample: Assign every member a number; draw numbers
      randomly using a computer generator or lottery bowl.
   2. Stratified Random Sample: Partition population into subgroups
      (e.g., 6th, 7th, 8th grades), then randomly sample proportionally from each.
   3. Systematic Random Sample: Select every kth person from an ordered list
      (e.g., every 10th student on an alphabetical roster after a random start).

Random sampling eliminates human bias from the selection process, providing a mathematically sound basis for drawing inferences about the broader population.

Exercises

Work through each problem carefully in your notebook before reading the provided solutions.

Exercise 1. A public library wishes to know the most popular book genres among town residents. The library places a paper survey on the front checkout counter for patrons to fill out voluntarily. Identify the population, the sample, and explain why this sampling method may produce biased results.

Exercise 2. A factory produces forty thousand light bulbs each day. To test quality, technicians test the first fifty light bulbs produced when the morning shift begins. Is this a representative random sample? Explain why or why not, and suggest a better sampling strategy.

Exercise 3. A school principal wants to survey student opinions on school lunch options. Explain how the principal can use an alphabetical directory of all eight hundred enrolled students to generate a simple random sample of eighty students.

Exercise 4. Contrast a voluntary response online poll on a news website with a scientifically designed telephone survey using random digit dialing. Which method produces valid statistical inferences, and why?

Solutions

Solution to Exercise 1. The population is all residents living in the town. The sample consists of the specific library patrons who choose to complete the survey at the checkout desk. This sampling method is biased because it relies on both a convenience sample and a voluntary response sample. It excludes all town residents who do not visit the library, and among library visitors, it only captures those motivated enough to fill out the form. The sample will heavily over-represent avid readers and library users while ignoring non-users.

Solution to Exercise 2. No, testing the first fifty light bulbs of the morning is not a representative random sample; it is a convenience sample. Light bulbs manufactured first thing in the morning may not reflect machine performance later in the day when equipment runs hot or components experience wear. A much better strategy is a systematic random sample: test one light bulb randomly selected every ten or fifteen minutes throughout the entire twenty-four-hour operating schedule, or assign batch numbers and use a random number generator to select fifty bulbs from across all manufacturing shifts.

Solution to Exercise 3. The principal can assign a unique identifying integer from one to eight hundred to each student on the alphabetical roster. Then, using an automated computer random number generator, generate eighty unique random integers between one and eight hundred (discarding any duplicate numbers). The eighty students whose roster numbers match the generated integers form the sample. Because every student on the roster had an equal one-in-eight-hundred chance of selection, this generates an unbiased simple random sample.

Solution to Exercise 4. An online poll on a news website relies on voluntary response sampling. Only readers who visit that specific website and feel strongly enough to vote will participate, typically over-representing individuals with extreme or polarized views. In contrast, a scientifically designed survey using random digit dialing reaches a randomized cross-section of households regardless of their personal motivation or political leaning. Therefore, only the randomized survey produces valid statistical inferences about the broader public.