STATA: An Introduction Into the Basics

Prof. Dr. Herbert Brücker

University of Bamberg

Seminar “Migration and the Labour Market”May 31, 2012

Repetition of last meeting (I/III)

1 The Borjas (QJE 2003) model

The regression equation:

yijt = wage or unemployment ratepijt = migration sharesi = vector of eduction dummies (i = 1, 2, 3, 4)xj = vector of work experience dummies (j = 1, 2, …, 8)πt = vector of time dummies (t = 1, 2, …, 25)si × xj = vector of time-experience dummiessi × πt = vector of education-time dummiesxi × πt = vector of experience-time dummies

Repetition of last meeting (II/III)

•The linear regression model• The univariate model• The multivariate model

•Econometric basics• endogenous (LHS) and exogenous (RHS) variables• regression coefficients/parameters• error (disturbance) term• assumptions of linear regression model• ordinary least squares estimator

•Interpretation of regression output• coefficient• standard deviation• t-statistics• p-value

Repetition of last meeting (III/III)

•Regression diagnostics• R-squared and adjusted R-squared• F-statistics• Degrees of freedom

Contents of Today’s Meeting

1The STATA Software package2The Structure of STATA: Three files3Getting started 4The STATA Menues5The General Structure of STATA6Working with DO FILES7Describe your data8Running regressions

1 STATA SOFTWARE PACKAGE

Image of STATA DVD in the Campus Net under:

“\\software\campliz”or:“\\software.uni-bamberg.de\campliz”

Then: Start -> AusführenThem: INSERT your licence:

Serial number: ….Code: ….Authorisation key: ….

2 Structure of STATA: Three files

1. The DATA file (.dta) where you have your data.• You can watch you data with the DATA

BROWSER and edit your data with the DATA EDITOR

2. The DO file (.do) where you run and save your commands of any session. Very useful (i) to organise your data set, (ii) to see what you have done in the last session, (iii) to replicate what you have done in last session, (iv) to exchange work with your collaborators. • You write and run your commands with the

DO FILE EDITOR

2 Structure of STATA: Three files

3. The LOG file (.log) which automatically reports all things which you have done during your session. Is automatically saved after your session. Not often used, but useful if something goes wrong.

3 Getting started: the STATA empty window

3 Getting started: The STATA empty window

The main window: shows commands, output and messages which arrive during your session

The command window: here you can type your commands

The variables window: Shows variables of your dataset

The review window reports your previous commands

3 Getting started: the windows after data loading

List of variables

Reports commands(one in this case)

Reports result of commands

3 Getting started

• In principle, you can start your STATA session by (i) loading your data set and (ii) typing your commands in the command window.

• It is however recommended to use the DO FILE EDITOR right from the beginning.

• But let’s look at the STATA menues first.

4 The STATA Menues

• For watching your data and changing your data by hand you need the DATA BROWSER and the DATA EDITOR.

• For starting and running your DO files you need the DO FILE EDITOR.

• The other menues are not relevant for the beginning.

The datapath

The dataeditor

The databrowser

The do fileeditor

The variablesmanager

The help menue

4 The STATA Menues: The DATA EDITOR/BROWSER

The difference between the data browser and the data editor is that you can manipulate data in the editor and only watch them in the browser.

4 The STATA Menues: The DATA EDITOR/BROWSER

You have two types of variables: NUMERICAL variables (black) and so-called STRING variables (blue) (e.g. text). STATA can identify STRING variables, but you cannot do numerical operations with them.

STRING variable

NUMERICAL variable

4 The STATA Menues: The DATA EDITOR/BROWSER

HINT: You can transfer data e.g. from an EXCEL file into a STATA file by copy and paste (STRG C + STRG V) and vice versa in the data editor. But you have to be careful that you EXCEL is run in English, otherwise your data might be read as STRING variables by STATA. Of course there are many other ways to transfer data from Excel to STATA.

5 The Grammar of STATA

General Structure of STATA commands

[prefix :] command [varlist] [if] [in] [weight] [, options]

5 General structure of STATA

We will concentrate on:

[prefix :] command [varlist] [if] [in] [weight] [, options]

5 General structure of STATA

We will concentrate on:

[prefix :] command [varlist] [if] [in] [weight] [, options]

What you want to do?

5 General structure of STATA

• There are two types of variables (data):• numerical variables, e.g.: 0, 1, 501, 0.5, -12 etc.• string variables, e.g.: no voc train , male, female etc.

• How to deal with the data types:• Numerical variables: you can do all mathematical operations,

e.g. var1 + var2, var1/var2, var1*var2 etc.

• String variables: You have to use quotation marks for identifcation, e.g.• var1 = 1 if sex == “female”

6 Working with DO FILES

• The standard approach is to start your work with a DO FILE

• Click on the DO FILE editor button after starting STATA

• Load an existing DO FILE or start a new one• Start the DO FILE with a command to load your data,

e.g.• use “path\data.dta”, clear

or, more specifically, with

• use “C:\Users\Herbert\Documents\STATA\Wagecurve\DE.dta", clear

Open your DO FILE editor

• After starting STATA click on the DO FILE editor button

The do fileeditor

How does a DO FILE look like

Commands

Descriptions of what you have done in stars *

The DO FILE menue

Clicking this button runs the entire DO FILE (not recommended)

Clicking this button runs a selection of marked commands (recommended)

Note: STATA stops the DO File execution after the first mistake in your commands. That makes it advisable to proceed step by step.

6 Step 1: Loading your data

• use “C:\Users\Herbert\Documents\STATA\Wagecurve\DE.dta", clear

• The use command loads the data• the “path\DE.dta” provides STATA the information

on the path where to find the data and the name of the data file (e.g. DE.dta)

• the clear command after the comma clears the memory, which is needed if you have used other data sets before

• Push the “Execute Selection (DO)” button to run the selected command(s)

• You can also run the entire DO File by pushing the “Execute Selection Quietly (RUN)” button

Loading your data (I/II)

1. Write the command use „path\XXX.dta“, clear2. Mark the line and run the command by clicking the

execution button

Loading your data (II/II)

6 Step 2: Manipulating your data (I/VI)

• It is useful to save only a basic data set and generate the variables you need at the beginning of each session. That saves storage space (recommended in case of large data sets)

• Generating DUMMY variables• Use the gen command, e.g.

• gen D_ed1 = 0• This creates a variable consisting only of zeros• Then use the replace command, e.g.

• replace D_ed1 = 1 if ed1 == 1 • This replaces the zeros with 1 if the variables ed1

has a values of 1.

Generating Dummy Variables: DO FILE commands

Generating Dummy Variables: STATA main window

6 Step 2: Manipulating your data (II/VI)

• Another example for generating dummy variables:• Use the gen command, e.g.

• gen year_1 = 0• This creates a variable consisting only of zeros• Then use the replace command, e.g.

• Year_1 = 1 if year == 1991 • This replaces the zeros with 1 if the year variable

has a values of 1991• Note: The STATA syntax requires that you have to

use after an if command always a double == for the definition of the value

6 Step 2: Manipulating your data (III/VI)

• Creating series of dummy variables if it is too cumbersome to create them individually, e.g. in case of interaction dummies

• Syntax:• forvalues i = 1/3 {

forvalues j = 1/4{ gen D_ed`i’*D_ex`j’

} }• i.e. for each value I = 1,2,3 and each value j =

1,2,3,4 you generate an interaction dummy by multiplying the dummy variables for education and experience. Take care of the {}!

Generating Dummy Variables: Advanced techniques

Generating Dummy Variables: Advanced techniques

Generating Dummy Variables: Advanced techniques

6 Step 2: Manipulating your data (IV/VI)

• Transforming variables into log variables• Syntax:

• gen ln_wijt = ln wijt• By using again the gen command you can transform

the wage variable wijt into the natural logarithm of the wage by applying the ln operator

Transforming data

6 Step 2: Manipulating your data (V/VI)

• Useful operators in STATA:

• + add• - subtract• * multiply• / divide• ln transform into natural log• exp transform into exponential value

6 Step 2: Manipulating your data (VI/VI)

• Control what you have done• Check you variables for mistakes in the browse

modus of the data set• You can delete wrong variable by using the drop

command, e.g.• drop ln_wijt

• Which simply drops your variable from the data set. Then you can create the correct one.

6 Step 3Organize your data with globals

• It is not convenient if you have to work with too many variables, e.g. 200 dummy variables (that is cumbersome to type some by hand)

• You can define globals, which comprise many variables

• Syntax:• glo [name of global [list of variables]• glo D_i Ded_1 Ded_2 D_ed3

• i.e the global D_i consists of the variables Ded_1 Ded_2 and Ded_3

• If you want to use the global later you have to type • $[globalname], i.e. $D_i

Creating globals

7 Describe your data (I/II)

• Any econometric analysis requires in the first step that you provide descriptive statistics to the reader. This helps to understand what’s going on

• This can be easily done with the sum command• sum [variable name(s)]• sum LHijt LFijt wijt ln_wijt

• The sum command creates a table with the complete descriptive statistics, i.e. observations, mean, standard deviation, minimum, maximum

Summary statistics

Summary statistics

7 Describe your data (II/II)

• Present your data graphically• It is usually helpful if you present the main

information /vairables in your data set graphically• There are many graphical commands, use the

Graphics menue• the simplest way is to show the development of your

variable(s) over time• Syntax:

• graph twoway line [variable1] [variable2] if …• graph twoway line wqjt year if ed==1 & ex == 1

• This produces a two-dimensional variable with the wage on the vertical and the year on the horizontal axis for education group 1 and experience group 1

Making a graph

Graph of mean wage in education 1 and experience 1 group

Graph of migration rate in edu 1 and exp 1 group

8 Running regressions

• The standard OLS regression command in STATA is• Syntax

• regress depvar [list of indepvar ] [if], [options] • regress ln_wijt m_ijt D_i D_j D_t

8 Running Regressions

Recall: What is a linear regression model

The general econometric model:

γi indicates the dependent (or: endogenous) variable

x1i,ki exogenous variable, explaining the independent variable

β0 constant or the y-axis intercept (if x = 0)

β1,2,k regression coefficient or parameter of regression

εi residual, disturbance term

Running a regression model

Regressioncommand

Dependentvariable

Independentvariables

Globals !

Running a Regression: Output

How to interpret the output of a regression

1. Observations2. fit of the model3. F-Test 4. R-squared5. adjusted R-squared6. Root Mean Standard Error

analysis of significance levels

variance of model

β0

β1

degrees offreedom

95% confidence interval

_cons 4.706176 .017403 270.42 0.000 4.672015 4.740337 mqkt -1.369118 .093913 -14.58 0.000 -1.553464 -1.184772 ln_wqkt Coef. Std. Err. t P>|t| [95% Conf. Interval]

Total 111.329246 799 .139335727 Root MSE = .33192 Adj R-squared = 0.2093 Residual 87.9145738 798 .110168639 R-squared = 0.2103 Model 23.4146717 1 23.4146717 Prob > F = 0.0000 F( 1, 798) = 212.53 Source SS df MS Number of obs = 800

. reg ln_wqkt mqkt

8 Running Regressions: Panel Models

• Very often you use panel models, i.e. models which have a group and time series dimension

• There exist special estimators for this, e.g. fixed or random effects models• A fixed effects model is a model where you have a

fixed (constant) effect for each individual/group. This is equivalent to a dummy variable for each group

• A random effects model is a model where you have a random effect for each individual

group, which is based on assumptions on the distribution of individual effects

8 Running Regressions: Panel Models

Preparation for Panel Models:• For running panel models STATA needs to identify

the group(individual) and time series dimension• Therefore you need an index for each group and an

index for each time period• Then use the tsset command to organize you dataset

as a panel data set• Syntax:

• tsset index year• where index is the group/individual index and year

the time index

Preparation: Running the tsset command

8 Running Regressions: Panel Models

• Then you can use panel estimatos, e.g. the xtreg estimator

• Syntax• xtregress depvar [list of indepvar ] [if], [options] • regress ln_wijt m_ijt, fe

• i.e. in the example we run a simple fixed effects panel regression model which is equivalent to include a dummy variable for each group (in this case education-experience group)

Running a Panel Regression: command

Running a Panel Regression: Output

Next Meeting:

June 14, 2012

Room RZ 00/006