This is synthetic data! While it's useful for getting a sense of the real data, it's not real data! It cannot be used to draw conclusions about Syntucky students or employed individuals.
To reiterate: this isn't real data. It mimics certain aspects of real data, but it was created by us using models that do not capture all aspects of the data generating process! It is not suitable for making inferences about Syntucky students or employed individuals.
We tried hard to protect individuals in Syntucky! For example, we deliberately chose methods for generating synthetic earnings data that do not include extreme outliers (consider why these values are particularly sensitive!). This results in distributions that deviate significantly especially from the upper percentiles.
We did preserve some logical dependencies from the original data. For instance, the variable representing the highest degree of a person will always have a higher value than the variable representing their first degree. We manually introduced these constraints in the synthesis procedure since the synthetic process we used is probabilistic and could produce "impossible combinations" of values.
There is a lot you can do with this dataset! Have fun! We have included a wide range of variables in the synthetic data and have put in considerable effort to implement privacy-safe models that still allow you to explore various measurements and cohorts. We hope it provides a realistic experience, but always remember that you are working with synthetic data!