UCLA Academic Technology Services HomeServicesClassesContactJobs
Search

Stata Code Fragments
Descriptives, ttests, Anova and Regression

* descriptive statistics:

summarize x1 
summarize x1, detail      /* 5 number summary */
stem x1                   /* stem & leaf diagram */
graph x1, normal          /* histogram */
lv x1                     /* letter-value table */
tabulate x1               /* one-way table */
tabulate x1 x2            /* two-way table */
table x1, by(x2) contents(mean y sd y)


* one sample:

ttest y = 10              /* single sample t-test */
signtest y = 10           /* sign test */

* two dependent groups:

ttest post = pre          /* dependent t-test */
signrank post = pre       /* Wilcoxen signed-rank test */

* two independent groups:

ttest y, by(grp)          /* independent t-test equal variances    */
ttest y, by(grp) unequal  /* independent t-test unequal variances  */
ranksum y, by(grp)        /* Wilcoxen rank-sum (Mann-Whitney) test */
hotel y1 y2 y3, by(grp)     /* Hotelling's T2                        */

* multiple groups:

oneway y grp, tabulate sidak   /* one-way anova */
anova y grp                    /* one-way anova */
kwallis drink, by(grtp)        /* Kruskal-Wallis test */

test a, error(a*b)


* regression: 

correlate y x1 x2 x3
regress y x1 x2 x3
logistic y x1
fit y x1 x2 x3
test x1
predict yhat               /* predicted score */
predict sresid, rstandard  /* standardized residuals */
rvfplot                    /* residual vs predicted */
rvpplot x1                 /* residual vs IV */
avplot                     /* added-variable plot (partial plot) */
avplots                    /* added-variable plots (partial plots) */
lvr2plot                   /* leverage to squared resid plot */
vif                        /* variance inflation factor */
correlate, _coef

* seemingly unrelated regression:


An economic model may contain a number of linear equations. It may be unrealistic
to expect that the equation errors will be uncorrelated. A set of equations that
has contemporaneous cross-equation error correlation is called a seemingly unrelated
regression (SUR) system. At first look the equations seem unrelated, but the
equations are related through the correlation in the errors.


use hsb2 
sureg (read gender ses ss)(math gender ses sci) 

* compare to regular regression
regress read gender ses ss 
regress math gender ses sci 


How to cite this page

Report an error on this page

UCLA Researchers are invited to our Statistical Consulting Services
We recommend others to our list of Other Resources for Statistical Computing Help
These pages are Copyrighted (c) by UCLA Academic Technology Services


The content of this web site should not be construed as an endorsement of any particular web site, book, or software product by the University of California