Topic 1.1 Introducing Statistics - What Can We Learn from Data?: identify questions to be answered, based on variation in one-variable data, and recognize what a data set can and cannot tell us.

What this topic is asking

The College Board (Topic 1.1) wants you to see statistics as the study of variation: data exist because individuals differ, and a statistical question is one that anticipates that variability and is answered by describing a group rather than stating a single fact. You must be able to identify what a one-variable data set can answer, and what it cannot.

Variation is the reason statistics exists

If every student were exactly $170\ \text{cm}$ tall, there would be no question to ask about height: the answer would just be $170$. Because heights differ, we get a distribution of values, and questions such as "what is a typical height?" and "how much do heights vary?" become meaningful. Throughout Unit 1 you describe that distribution using three lenses: shape, center, and spread.

What a data set can answer

A one-variable data set lets you answer descriptive questions about the group measured:

• Center. What is a typical value? (mean, median)
• Spread. How much do the values vary? (range, IQR, standard deviation)
• Shape. Is the distribution symmetric, skewed, or does it have clusters, gaps, or outliers?
• Comparison within the data. How does one subgroup compare with another, if the data are grouped?

What a data set cannot answer

Recognizing the limits is half the topic. A single observational one-variable data set cannot:

• Establish causation. Observational data can reveal an association, but cause requires a designed experiment (Units 3 to 4 territory).
• Predict an exact future value. Data describe what happened; they do not guarantee any individual outcome.
• Generalize beyond the group unless the data came from a proper random sample of a defined population.
• Answer a non-statistical (deterministic) question about a single known individual, which is just a lookup, not statistics.

Statistical versus deterministic questions

The cleanest way to tell whether a question is statistical is to ask whether the answer would vary if you collected the data again, or whether it is a fixed fact. "How tall is the tallest student in this room right now?" is deterministic: there is one correct answer and no variability to anticipate. "How do the heights of students in this room vary, and what is typical?" is statistical: it anticipates that the values differ and is answered by summarizing the spread and center. The College Board threads this distinction through the whole course. Hypothesis tests later in the course are simply formal statistical questions about a population, asked when variation in a sample leaves room for doubt. So the habit you build here, of checking "does this question anticipate variability, and is it answered by a distribution?", is the same habit that lets you choose the right inference procedure much later. A question that asks only about one fixed individual, or that asks for an exact prediction, falls outside what data can deliver, and saying so plainly earns marks on the exam.

Why scope matters on the exam

Free-response questions in this course frequently ask you to interpret in context and to state limitations. Topic 1.1 is where you first practice that discipline: when handed a data set, you should be able to name a question it answers and, just as importantly, name a question it does not, with a reason. The commonest reason a data set falls short is that the question is causal while the data are merely observational, or that the question concerns a variable that was never measured. A second common reason is generalization: a convenience sample (for example, only the students in one class) cannot support a claim about all students in the school, because the sample was not random. Training yourself to spot these gaps now means that when a later free-response question says "explain why the researcher cannot conclude that X causes Y," you already have the language ready, and you will not be tempted to over-claim from the data in front of you.

Try this

Q1. Identify one statistical question that can be answered by a data set listing the daily rainfall (mm) at one weather station for each day of a year. [1 point]

• Cue. Any question about the distribution, for example "what is a typical daily rainfall, and how much does it vary across the year?"

Q2. Explain why "How much do these values vary?" is a statistical question but "What is the value for individual 7?" is not. [2 points]

• Cue. The first anticipates variation across the group and is answered by describing spread; the second is a deterministic lookup of a single fixed value with no variability to summarize.