Using STATA in PS400

STATA is a fairly straightforward statistical package that has a wide range of statistical procedures.


It also has some useful support on line on their WebSite.

This latter document makes a good tutorial.

This Web page providees a few basis:


First, find and execute stata (In other words, either find Stata in the Start > Program hierarchy, or double click the STATA icon)

This brings up the Stata "desktop."

The desktop has several components

In addition, Stata has some other Windows you will need to use:

The Menu Bar

Like most windows programs, Stata uses a set of menu items from the top bar of the window. Run through them looking at the options available to you. Note especially the text label that appears if you pause the mouse over a button on the bar.

The Command Window

You run Stata by typing cpommands in the Command window. Needless to say the first command you will likely enter is one to open a data set:


For instance, to open up thenclass data set on presidential approval, type

use "E:\WWW\DUVAL\ps400\Notes\Stata\presapp.dta", clear

Note that this command can be executed with the File > Open menu item in the Menu Bar as well

Also examine the Stata Results, Variables, and Review Windows after executing this command


The Stata Results Window



The Variables Window



The Review Window








Some Useful commands:

There is a basic subset of comands that are useful in using Stata. In general they fall into four groups:

File commands Data manipulation Statistical Commands Misc Utilities
  • use
  • insheet
  • generate
  • replace
  • sort
  • regress
  • dwstat
  • prais
  • sw


There is a data set available to explore these commands, and several commands listed below have examples that use this data set.. Try using each command and seeing how Stata responds when you use it. The data set is: presapp.dta (Stata Data set)

Type Command Description
File Commands  
  use

Open a Stata data set

e.g. open e:\www\duval\ps401\notes\spreadsheets\presapp.dta

  insheet

Imports a text delimeted file saved from a spreadsheet

e.g. insheet using e:\www\duval\ps401\notes\spreadsheets\presapp.txt

  describe Provides a list description of the variables in the data file.
  tsset

Defines a data set as being a time series data set

tsset time

time is the variable with unique indicator of temporal order. (year, or t, etc - not day, week, month or quarter)

Data manipulation  
     
  sort Sort the data
  gen

Generate a new variable

gen x3 = x1 + x2

  replace

Replaces an already existi ng variable wirth new information

replace x3 = x1 + x2

replaces the earlier version of x3

Statistical commands  
  corr

Produces a Pearsons correlation matrix

corr y x1-x5

produces the matrix for all six variables

  regress

Regression analysis

reg y x1-x3

regresses y on x1, x2, and x3

  dwstat calculates a Durbin-Watson statistic for the previous regression model
  prais

Runs a correction for autocorrelation

prais y x1 x2, corc

the corc option specifieds the Cochran-Orcutt method

Miscellaneous Utilities  
  sw

Stepwise - wraps a command in a stepwise iteration loop - appears only final step available.

sw reg y x1 x2 x3, pr(.2),pr(.05)

runs a tepwise regression with a p<.05 needed for inclusion and a p>.20 needed for removal

Stata commands are executed by typing them in the STATA Command text box and then hitting return. For instance to describe the data in the open data set, type

describe

As shown

which produces the following resuls in the STATA resulst box.


Useful Statistical Procedures