Who or What Is the Question About?
In Writing Investigative Questions That Data Can Answer, you practiced identifying questions that could be answered with relevant data and that anticipate variation. The next step is to say clearly who or what those data should describe. A question about “students,” “customers,” or “rainfall” may sound specific, but often leaves important boundaries unstated.
The group or set of observations a statistical question is intended to describe is called its population of interest. Defining this population makes the scope of the question clear before anyone decides how to collect data. It also helps prevent a conclusion about one group, place, or period from being presented as if it applied to a different one.
A population does not have to consist of people. It might be all bicycles checked at a repair shop during a month, all deliveries made in a town during a week, or all hourly temperature readings at specified locations over a season. The key is to identify the units that could each contribute an observation to the question.
That last point matters when people and measurements could be confused. If a question concerns students’ daily travel times, the units might be students, with one reported travel time per student. If it concerns trips, a student who travels on several days could contribute several observations, and the units might instead be trips. State which one the question is about.
Make the Boundaries Explicit
A practical way to define a population is to check three boundaries: who or what is included, where the units are located or observed, and when they qualify. These boundaries should match the question’s purpose, not simply whatever data happen to be easy to get.
Say whether the population consists of people, objects, events, or measurements. If it consists of people, identify which people; if it consists of events or measurements, define what counts as one unit.
Identify the relevant school, workplace, neighborhood, service area, or measurement site. Use a boundary precise enough that someone could tell whether a unit belongs.
State the relevant date, interval, season, or other period. If the question concerns events or repeated measurements, clarify which occurrences are included.
Confirm that the variable and the intended conclusion concern this exact population—not a broader or different group.
A time boundary can describe when people qualify for a population or when an event happened. For example, “students enrolled at a school on October 1” defines membership at a point in time. “Bus rides taken during October” defines a set of events over an interval. A question about a student’s sleep last night has a different time scope from a question about that student’s usual sleep on school nights.
A location boundary works the same way. “People in the area” is difficult to interpret unless the area is named or otherwise defined. “Visitors to Northside Park” is clearer if the park is identified, but the question may still need a time period and a rule for which visits count. Boundaries should be detailed enough for the intended question without adding irrelevant qualifications.
The Population Is Not Automatically the Data You Have
The population of interest is the group the question aims to describe. The people or units from which data are actually collected may be only part of that group. For example, a question may be about every member of a school’s debate club, while the only available responses are from members who attended one meeting. The intended population and the observed group are not the same description.
This distinction does not, by itself, establish whether collected data represent the population well. It does make the intended scope visible, so the question and eventual conclusion can be stated accurately. If data come only from a particular set of respondents or observations, do not silently redefine the original population to make it appear that the question was broader or narrower than it was.
This connects to the question audit in Writing Investigative Questions That Data Can Answer. There, you identified the subject and variable and checked whether data could answer the question. Now make the subject’s boundaries explicit. A question such as “How long do deliveries take?” identifies a variable, but it does not yet identify which deliveries, where they occur, or when they are made.
Worked Example: Define a School Population
Worked Example: Define a School Population
A student council asks, “How much time do students spend getting to school?” The question anticipates different travel times, but “students” could refer to several schools, and “getting to school” could mean a typical commute or a trip on a particular day.
Choose the units and boundaries. Suppose the council wants to describe travel times for students currently enrolled at Cedar Valley High School. It defines the variable as each student’s reported one-way travel time on the morning of October 6, measured in minutes. The location boundary is the named school’s enrollment, and the time boundary is that morning’s trip.
A clearer investigative question is: “What are the one-way travel times, in minutes, for students enrolled at Cedar Valley High School who traveled to school on the morning of October 6?” Here, the observational units are students who made that trip, and each contributes a travel-time observation. Students who did not travel to school that morning are outside this particular population as defined.
The council might instead want to ask about typical travel time across the school year. That requires a different time definition, such as travel times for enrolled students on school mornings during a stated month. It might also require clarifying whether the unit is each student or each student-morning. A single student could then have multiple trips in the data.
Conclusion. The original question gave a variable but not a complete population. Naming the school and the date makes its scope clear; choosing a longer period would define a different population and potentially a different unit of observation.
Worked Example: Define a Population of Measurements
Worked Example: Define a Population of Measurements
A community group asks, “How muddy is the river?” The wording does not specify where or when the river should be measured, and “muddy” does not name a precise variable. The group decides to study turbidity, a numerical measure recorded by a water-testing instrument.
Define the units. The group will use one turbidity reading as an observational unit. It selects three named monitoring points along the river and plans one reading at each point on each Monday in June. The population of interest is all those planned readings: readings at the three specified points on the Mondays in June.
The group can now ask: “What is the distribution of turbidity readings at the three designated river-monitoring points on Mondays in June?” This question identifies the variable, the locations, and the time period. Its units are readings, not the river as a whole and not individual water molecules.
If the group wants to describe turbidity throughout the entire river or throughout the whole year, this population definition would not cover that broader scope. It would need to revise the locations, time period, or both. Likewise, readings at an additional site would not belong to the population as currently defined.
Conclusion. For questions about measurements, define where and when each measurement is taken. “The river” is too broad to specify which readings are relevant; the named sites and Mondays in June create a clear population of observations.
Worked Example: Define a Population of Events
Worked Example: Define a Population of Events
A neighborhood delivery service wants to investigate how long deliveries take. Its first question is, “What is the usual delivery time?” The question does not identify which deliveries, where they are made, or what “usual” means in terms of the observations being studied.
Specify which events count. Suppose the service is interested in completed deliveries made within the town of Pine Ridge during September. It defines one delivery as one order completed at a customer’s address and measures delivery time from the time the order leaves the service’s preparation area to the time it reaches that address.
The population of interest is all completed deliveries meeting that definition within Pine Ridge during September. A suitable investigative question is: “What is the distribution of delivery times, in minutes, for completed orders delivered within Pine Ridge during September?” The units are delivery events. If one customer receives two orders, those are two events under this definition.
An order that is canceled before delivery is not part of this population, because it was not completed. A delivery made outside Pine Ridge is also excluded, even if the service made it during September. Stating the inclusion rules makes these decisions consistent rather than leaving them to guesswork.
Conclusion. A population of interest can be defined by events as well as by people. The event, location, time interval, and operational rule for what counts as one delivery make the question’s intended scope explicit.
Common Mistakes and AP Exam Tips
- Using a broad group name without boundaries. “Students,” “residents,” or “customers” may cover multiple places or groups. A full-credit response identifies which students, residents, or customers the question concerns.
- Leaving out the time period. “How much do residents recycle?” could refer to a day, month, or year. State a period that matches the question, such as recycling during a specified month.
- Confusing the unit with the variable. “Travel time” is a variable, not a population. Name the units that have travel times, such as students’ trips to school on a particular date.
- Mixing up people and repeated observations. A question about students and a question about student-days do not necessarily concern the same units. State whether each person contributes one observation or whether repeated occasions are included.
- Calling available respondents the intended population. The people who answer a survey may not be the full group the question is about. Identify the intended population separately from the group that actually provides data.
- Claiming a broader scope than the definition supports. Readings from three sites in June describe the specified readings, not automatically every location or season. Keep conclusions within the defined boundaries.
For a clear AP Statistics response, name the population in enough detail to tell who or what belongs, identify the location, and give the relevant time period. If the units are events or repeated measurements, say what counts as one unit. Then check that the investigative question and variable match that population.
Check Your Understanding
For each question, identify what additional population boundaries or unit details would make its scope clear.
- “How long do students spend on homework?” Give an example of a more clearly bounded population, including a location and time period.
- A question concerns the number of bicycles repaired at a shop. What could count as one observational unit, and what location and time boundaries could be specified?
- A group records air-quality readings at two locations every afternoon in April. Are the units people, locations, or readings? Explain what population of observations is defined.
- A delivery question concerns completed orders delivered within a town during one month. Name one event that would be excluded under that definition and explain why.
- Explain the difference between a population of interest and the group of people or units that actually provide data.