Predicting Population Outcomes
Introduction
Welcome to lesson six of Drawing Conclusions from a Sample — we are nearly at the finish line! By now you have built a strong toolkit: estimating population means and proportions from random samples, understanding that different samples produce different estimates, and gauging the likely error of any single estimate with a plausible range. Today we put all of those skills to work for their most practical purpose — making a prediction about an entire population based on what a sample tells us.
From Estimates to Predictions
So far in this course, we have focused on producing numbers: a sample mean, a sample proportion, a plausible range. Those numbers are valuable on their own, but in many real situations the ultimate goal is not just a number — it is a claim or a decision. A city council does not just want to know that 54% of surveyed residents favor a new park design. They want to know whether the whole city likely favors it. A product team does not just want to know that sampled customers rate Feature A higher on average. They want to predict which feature the full customer base prefers.
This is the natural final step in working with samples. You collect data from a randomly chosen subset, compute an estimate, and then extend that finding to the entire population as a prediction. The prediction is a statement about the larger group — not just the people you happened to survey — and it is justified precisely because the sample was selected at random.
Reporting a Sample Finding vs. Making a Population Prediction
Before we jump into examples, let's clarify an important distinction. A sample finding describes what happened in the data we collected. A population prediction uses that finding to make a claim about the larger group. These two statements may sound similar, but they serve very different purposes.
Consider a random sample of 200 customers asked whether they prefer Product A or Product B. Suppose 118 chose Product A and 82 chose Product B. Here are two ways to communicate that result:
- Sample finding: "In our sample of 200 customers, 59% preferred Product A."
- Population prediction: "We predict that Product A is preferred by the majority of all customers."
The first statement is safe and accurate, but it only describes the 200 people we happened to survey. The second statement reaches beyond the sample to make a claim about the entire population. That reach is exactly what makes predictions useful — and it is justified when our sample was selected randomly and is large enough to trust.
In this lesson, we will practice that move using examples where the sample lead is fairly clear. When the sample results are nearly tied, we need an extra uncertainty check before making a confident prediction — that close-call judgment is the focus of the final lesson.
