Two Page Stata

An Introduction to Stata

1 Introduction

An introduction to Stata in 2 pages.¹ Commands that you actually type into Stata are represented in monospace font. x and y refer to variables in your data. The treatment here is intended to be extremely brief, in order to create a kind of “cheat sheet” that can be presented in 2 pages. More documentation on any command is available in the printed or PDF Stata manuals, or by typing help command.

So many projects have the same, or similar, workflow.

A common data workflow

Tip

A little bit of Stata can go a long way.

The general idea of most Stata commands is command variable(s), options. Often it is not necessary to use any options since the authors of Stata have done such a good job of thinking about the defaults.

The Stata interface makes it extremely easy to do rapid interactive data analysis. Hit PAGE-UP to recall the most recent command, which you can then quickly edit and resubmit.

Use the DO FILE EDITOR to save Stata commands that you want to use again in a .do file, and to create an audit trail of your work so that your workflow is documented and replicable.

2 Get (And Get Acquainted With) Data

Once you are in the right directory, use "mydata.dta" will open your data. save "mydata.dta", replace will save your data.

It is good practice to start by looking at your data. browse x y z will open a data window with variables x, y and z.

codebook x y will produce a nicely formatted codebook of selected variables, which is especially useful if you have added variable labels and value labels. codebook is especially useful for seeing how numerical values are associated with value labels. codebook by itself will list every variable in your data and generate a lot of [probably too much] output.

lookfor allows you to find variables that contain a specified keyword. This is especially useful in large data sets with many variables. Often abbreviated keywords are the most helpful. e.g. to find a poverty variable, type lookfor pov.

With very large data sets, it may be helpful to use keep x y z to only keep the variables with which you are working.

describe tells you about the contents of a specific variable. E.g. describe x y. describe, short will tell you very basic things about your data, including the number of observations in the data set, and the size of your data file.

3 Process And Clean Data

Data with missing values, often represented as negative numbers (e.g. -99, -9, -8) need to be recoded so that the missing values are represented as a missing value character (“.”) that Stata knows to exclude from calculations.

recode x (oldvalue = newvalue), generate(xR)² will recode a variable into a new variable. recode x (-99/-1 = .), generate(xR) will recode negative numbers from -99 to -1 to missing for x. recode x (7/9 = .), generate(xR) changes 7 through 9 to be missing for x. Indeed, recode will change specific values in your data to anything you want, not just missing values. Reverse coding often looks something like recode x (1=3) (2=2) (3=1), generate(xR).

Like many other statistical programs, Stata makes a large distinction between variables that are coded as numeric, and variables that are coded as strings. describe x will help you to ascertain the variable type. encode x, generate(x_NUMERIC) is often useful to create a numeric version of string variables. In the special case where the values of your variables are actually numbers but stored as strings, destring x, generate(x_NUMERIC) may be the more helpful and appropriate command. The differences between the two commands are subtle, so consult Stata help on each command.

It is often convenient to rename your variables so that the variables have more intuitively understandable names e.g. rename x depression.

You can create new variables out of old variables using generate newvar = expression e.g. generate newvar = oldvar1 + oldvar2.³

It is sometimes useful to sort your data. sort x will sort your data by the values of x.

4 Analyze Data

4.1 Descriptive Statistics

summarize gives you basic descriptive statistics for a variable, such as the mean (average), especially useful for continuous variables. E.g. summarize x y or summarize x y, detail. tabulate gives you a frequency distribution for your variable, especially useful for categorical variables. e.g. tabulate x.

4.2 Bivariate Statistics

Tabulating two categorical variables together gives you a cross-tabulation of those variables, e.g tabulate x y, row col chi2. pwcorr x y, sig gives you the pairwise correlation of two continuous variables. oneway x z, tabulate gives you a oneway ANOVA of continuous variable x over categorical variable z.

4.3 Multivariate Statistics

regress y x regresses y on x.⁴ regress y x z regresses y on x and z.⁵ regress y x i.z regresses y on x and z, treating x as continuous and z as a set of categorical indicator variables.⁶ regress y c.x##i.z regresses y on continuous x and categorical z, providing both main effects for x and z and the interaction of x and z.

5 Visualize Data ^{7 8}

histogram x will give you a nice display of one variable.⁹

twoway scatter y x gives you a scatterplot of your data. twoway lfit y x will give you a linear fit graph. The two syntaxes may be combined e.g. twoway (scatter y x) (lfit y x).

graph bar, over(x) is useful for creating a bar graph of the counts of a categorical variable x. graph bar y, over(x) will create a bar graph of the means of y over categories of x.

Footnotes

1. Comments, questions and corrections most welcome and may be sent to: Andrew Grogan-Kaylor @ agrogan@umich.edu. This document available on the web @ https://agrogan1.github.io/Stata/

2. While generating a new variable is optional, it is almost always a good idea.

3. alpha oldvar1 oldvar2 will calculate Cronbach’s alpha from this scale.

4. After running a multivariate model estat summarize will give you simple descriptive statistics for the specific sample used in that particular analysis.

5. Other regression commands follow a very similar format: command y x z but are beyond the purview of this 2 page guide.

6. i.x is Stata’s notation for treating independent variables as categorical or indicator variables.

7. For all graphs, options after a “,” will be helpful in titling your graph e.g. twoway lfit y x, title("...") xtitle("...") ytitle("...")

8. Graph schemes can change the overall look of a graph. , scheme(s1color) is often a good choice. In newer versions of Stata, , scheme(stcolor) works well.

9. histogram x, percent will scale the y-axis more intuitively in terms of percentages. histogram x, discrete gives a nicer display for categorical variables. The percent and discrete options can be combined.