# Four Methods of Collecting Data

Analyzing data can help us answer important questions. But before we can analyze data, we first have to collect it.

There are four methods we can use to collect data.

## Method 1: Census

Definition: A census is a study that collects data for every member of a population. In most cases, a census is not used because it’s too time-consuming and expensive to collect data for every member of a population.

Example: We want to know the average height of a professional basketball player so we measure the height of every single player in the league. This is a census since we collect data for every member of the population we’re interested in.

## Method 2: Sample Survey

Definition: A sample survey is a study that collects data on a subset of the population, which is then used to draw inferences about the larger population.

Example: We want to know how many citizens in a city with a population of 50,000 are in favor of passing a new law. We randomly select 500 citizens and ask them whether or not they are in favor of the new law. We use the responses from these 500 citizens to make an inference about what percentage of the total population would support the new law.

## Method 3: Experiment

Definition: An experiment is a study in which a researcher assigns subjects to a specific group  and then decides what type of treatment that group receives. The purpose of an experiment is to understand a cause-and-effect relationship between a treatment and an outcome.

Example: A researcher wants to know if a new diet causes weight loss. The researcher splits up 50 people into two groups of 25. One group is told to follow the new diet for one month. The other group is told to just eat what they normally do. At the end of one month, the researcher will analyze the weights of the people in each group to determine if the new diet does or does not cause weight loss.

## Method 4: Observational Study

Definition: An observational study is a study that also attempts to understand a cause-and-effect relationship between a treatment and an outcome. But unlike experiments, researchers can only “observe” subjects in an observational study. Researchers have no ability to assign subjects to specific groups or decide what type of treatment they receive.

Example: A researcher wants to know if smoking impacts overall health. They find 100 people, 50 of whom have not smoked during the past five years, and 50 of whom have smoked every day for the past five years. The researcher analyzes their health and draws conclusions about whether or not smoking impacts overall health. This is an observational study because the researcher didn’t do anything except observe the people in the study without influencing them in any way.