First you’ll have to understand the two other main types of data, cross sectional and time series. Cross sectional data looks at a snapshot in time, and compares a lot of different attributes of the given subjects in a sample as they exist at that particular time. It generally compares differences among the subjects. Think of the test scores of a particular class of kindergartners, and how they vary according to the differences among the kids, their parents, etc.
Time series data tracks a given variable/measure over time. Think of a child’s test scores as he ages.
Panel data is a mix of both. It starts as a cross sectional data set, by taking a sample of people and measuring some number of their attributes at some time. But then it regularly “checks back in” on the *same* sample, by re-taking all those measurements at regular intervals, say annually. So in our example, we sample the kindergartners, then re-do the sampling each year, all the way through high school.
First you’ll have to understand the two other main types of data, cross sectional and time series. Cross sectional data looks at a snapshot in time, and compares a lot of different attributes of the given subjects in a sample as they exist at that particular time. It generally compares differences among the subjects. Think of the test scores of a particular class of kindergartners, and how they vary according to the differences among the kids, their parents, etc.
Time series data tracks a given variable/measure over time. Think of a child’s test scores as he ages.
Panel data is a mix of both. It starts as a cross sectional data set, by taking a sample of people and measuring some number of their attributes at some time. But then it regularly “checks back in” on the *same* sample, by re-taking all those measurements at regular intervals, say annually. So in our example, we sample the kindergartners, then re-do the sampling each year, all the way through high school.
It adds a time element to cross-sectional data.