Introduction
A broad topic such as student study habits does not yet tell us what to record or what question the data should answer. A descriptive question narrows the topic to a single variable measured on a defined group. For example, instead of asking generally about study habits, we might ask how many hours students in a specified class year spent studying outside class during a specified week.
The aim is not to predict what the answer will be. It is to ask clearly enough that each individual can contribute a value and the collection of values can be described. Those values may differ: some students may report few study hours and others many. Describing that spread or pattern is the point of the question.
From a Topic to a Question
A topic names an area of interest; a descriptive question identifies what information is needed. “Campus transportation” is a topic. “How long did each student’s trip to campus take on the most recent day they attended an in-person class?” is a question about one variable: commute duration. The question directs what to record for each student and makes the answers available for description.
A useful question usually includes three parts:
- Group: Which individuals are described? The group should be specific enough that readers know who belongs in it.
- Variable: What one characteristic will be recorded for each individual? Name it in terms that indicate what counts as a value.
- Time or setting: When, or under what conditions, is the value recorded? Include this when values could change with time or context.
For instance, “How much do students study?” leaves all three parts open. Which students? What counts as studying? Over what period? A more answerable version might specify undergraduates enrolled in a particular course, total hours spent studying outside scheduled class meetings, and the previous seven days. The detail is not decoration: different choices can produce different data.
The variable can be categorical or numerical. A question about class year asks for categories such as first-year or second-year. A question about weekly study hours asks for a numerical amount. In either case, the question concerns one variable recorded for individuals in a defined group.
Worked Example: Weekly Study Hours
Suppose a class instructor wants a description of study time among the six students in a hypothetical seminar during one particular week. The records for total hours spent studying outside scheduled class meetings are 4, 6, 7, 8, 10, and 11 hours.
A focused question is: How many hours did each of the six students in the seminar spend studying outside scheduled class meetings during the week of October 5–11? The group is the six seminar students, the variable is weekly hours of out-of-class study, and the time period is specified.
The values are numerical and are not all identical. For this small hypothetical set, the smallest recorded value is 4 hours and the largest is 11 hours. Their difference is:
This calculation illustrates that the recorded values vary; it does not change the question’s focus on one variable. The data describe these six students for this week only. They do not, by themselves, establish what students in other seminars or weeks typically study.
What Makes the Question Statistical?
A question is statistical when answering it requires data that can vary across individuals, rather than looking up one fixed value. If we ask, “How many students are enrolled in this seminar?” the answer is one count for the seminar. If we ask, “How many hours did each student study this week?” the answer consists of values that may differ from student to student. The second question calls for a description of a variable across individuals.
The distinction is about what the question asks us to collect, not whether the values are guaranteed to be different. A group might happen to have identical values. The question is still statistical if it asks for the value on each individual and is designed to describe how those values are distributed. Before data are collected, the variation is possible rather than guaranteed.
A short way to check the structure is to imagine one row for each individual and one column for the variable. Can you say what value belongs in that column for each person? If not, the variable may still be vague, or the group or time frame may need to be specified. This check helps turn an interesting topic into a question with a clear answer in data.
Worked Example: Class Year in a Defined Group
Imagine a hypothetical set of 20 students enrolled in an undergraduate biology course on October 1. Each student reports their current class year. The recorded category counts are 5 first-year students, 6 second-year students, 5 third-year students, and 4 fourth-year students.
A descriptive question is: How are the 20 students enrolled in the biology course on October 1 distributed across class year? The group is the 20 course enrollees on that date, and the single variable is class year. Since the values are categories, an appropriate description reports how many students fall into each category.
The counts account for the whole group because:
For example, the proportion recorded as second-year students is \(6/20 = 0.30\), or 30%. This is a description of the hypothetical group, not a claim about all biology students. The question concerns the distribution of one categorical variable; it does not require assigning numerical meaning to the class-year labels.
Make the Measurement Clear
A single-variable question is only as clear as its variable. Everyday terms can refer to different measurements. “Commute” might mean distance, duration, number of travel modes, or a category such as walking or taking a bus. A question should make clear which one value is recorded for each individual.
Time-sensitive measurements need a time reference. “How long is a student’s commute?” could refer to an average across many trips, a usual trip, or one particular trip. If the intended variable is the duration of one trip, state which trip or occasion. This matters even when the group remains unchanged, because one person can have different values on different days.
The question also needs to define its group in usable terms. “Students” could mean all students at a university, students in one course, or students living on campus. A defined group limits what the question describes. It should name the relevant membership and, where appropriate, the date or period that determines who is included.
Worked Example: Commute Duration
Consider six hypothetical students. For each, a recorder notes the number of minutes from leaving home to arriving on campus on the most recent day the student attended an in-person class. Their recorded durations are 12, 18, 21, 21, 38, and 45 minutes.
A focused question is: How long did each of the six students take to travel from home to campus on the most recent day they attended an in-person class? The group consists of these six students, the variable is commute duration in minutes, and the stated occasion identifies which trip to record.
The recorded values range from 12 to 45 minutes. The difference between the endpoints is:
This gives a simple indication of how widely the hypothetical durations vary. It does not explain why one trip took longer than another. The question asks what durations were recorded, not what caused them.
A Brief Justification: Why These Details Matter
Suppose a group contains several individuals and a question asks for the value of one variable for each. Recording the answers gives one collection of values. That collection can be described because its members may differ: categories may occur at different frequencies, or numerical values may span different amounts. If the question omits the group, we cannot tell whose values belong in the collection. If it omits the variable’s meaning, we cannot tell what each value represents. If it omits a needed time or setting, two people may record different kinds of occasions.
Thus the three details work together. The group determines whose values are in scope, the variable determines what is recorded, and the time or setting determines the circumstances of the measurement. Once those are specified, the question can be answered by collecting values and describing their pattern. This is why a well-formed descriptive question is more than a topic label.
Common Pitfalls
- Leaving the group broad. “How many hours do students study?” does not say which students. Name the course, campus, or other membership that defines the intended group.
- Using a vague variable name. “How much do students commute?” could refer to several measurements. Specify duration, distance, or another single measure, with units where relevant.
- Leaving out the relevant time. “How long is the trip?” may produce different answers depending on the trip. Identify the occasion or time period when values can change.
- Asking for an explanation instead of a description. “Why do some students study more?” asks about causes, not just the values of one variable. A descriptive question first asks what study hours were recorded for the defined group.
- Assuming the answer before collecting data. A question such as “How many hours do students waste studying?” builds a judgment into its wording. Ask neutrally what hours were recorded; the description should follow the data.
Check Your Understanding
For each question, identify what is clear and what should be specified or revised.
- “How much do students study?” Rewrite this as a question with a defined group, one measurable variable, and a time period.
- A researcher records the class year of every student in a specified course on September 15. What is the single variable, and what kind of values does it take?
- Why might “How long is the commute?” lead to inconsistent answers? Revise it to identify a specific trip or time reference.
- Explain why “How many students attend this lecture?” is different from asking how many hours each student studied during the previous week.
- A question asks, “Why are some students’ commutes longer?” Describe how to turn it into a descriptive question about one variable for a defined group.