Stata: Working with Variables

Useful tutorials on Stata and other stats software.

Guide from UCLA Stats Consulting, with chart of tests by number and type of independent and dependent variables.

List of educational resources compiled by the Stata Corp.

Tutorial from the Carolina Population Center, including tutorials for complex sample survey data.

How to load ICPSR data into Stata.

Menu for importing data from FRED. An API key can be requested.

App that allows you to access hundreds of free time series directly from within Stata.

App for importing over 3,000 series of indicators from the World Bank databases directly into Stata.

encode varname, generate(newvar) will convert a categorical string variable into a numeric one with labels.

destring varlist, generate(newvarlist) is what you want when your string really is a number (often happens when you import data with dollar signs or NA values).

To create dummy variables from variable group, you may use tab group, gen(g), or use a factor variable, e.g. regress x y i.group

A little bit about what dummy variables are for.

Why you shouldn't be alarmed if one of your dummy variables disappears.

Probability weight? Frequency weight? What? Find out the difference.

Most Stata commands can include your weight variable. This help page provides some examples.

Working paper discussing whether to use weights in an analysis.

Install using "ssc install cpigen." Adds a consumer price index variable to your dataset.

If cpigen doesn't have the years or series that you want, you can also get CPI data from FRED. Either download from the FRED site or go to File > Import > Federal Reserve Economic Data, and search for CPIAUCSL.

Before using a time series analysis, tsset tells Stata how your time variable is stored.