/* SAS Program */

options nocenter;



data hsb200;
  infile 'hsb200.txt';
  input id gender $ race ses
        schtype $ prog read write
		math science socst;
        
  /* convert string variable to numeric */      
  sex = .;
  if gender = 'm' then sex = 2;
  if gender = 'f' then sex = 1;

  /* create dichotomous dummy variable */
  /* assumes all write scores are non-missing*/
  honor = (write ge 60);


  /* create difference variable 
     for paired t-test */
  rminusw = read - write;
run;

/* descriptive statistics */
proc means data=hsb200;
  var read write math;
  
/* exploratory data analysis */
proc univariate data=hsb200 PLOT;  
  var write;
  



/* frequency tables */
proc freq data=hsb200;
  tables gender race ses;
  
/* cross tabulation */
proc freq data=hsb200;
  table gender*race / chisq;
  
/* correlations */
proc corr data=hsb200;
  var read write math sex;
  
/* paired t-test */
proc means n mean t
     stderr prt data=hsb200;
  var rminusw;
  
/* independent t-test */
proc ttest data=hsb200;
  class gender;
  var write;
  
/* multiple regression */
proc reg data=hsb200;
  model write = sex read math science;
quit;  
  
/* logistic regression */
proc logistic data=hsb200 descending;
  model honor = sex read math science;


/* factorial anova */
proc glm data=hsb200;
  class gender prog;
  model write = gender prog gender*prog;
quit;  
run;

/* Stata Program */



infile id str1 gender race ses /*
    */ str3 schtype prog read write /*
    */ math science socst   /*
    */ using hsb200.txt


/* convert string variable to numeric */      
encode gender, generate(sex)
    
    
/* create dichotomous dummy variable */
/* assumes all write scores are non-missing*/
generate honor = (write >= 60)


/* do not need to create difference  
variable for paired t-test */
  
  

/* descriptive statistics */
summarize read write math
  
  
/* exploratory data analysis */
summarize write, detail  
stem write
graph write, box
pnorm write  


/* frequency tables */
tab1 gender race ses


/* cross tabulation */
tabulate gender race, chi2
  
  
/* correlations */
correlate read write math sex
  
  
/* paired t-test */
ttest read = write
  


/* independent t-test */
ttest write, by(gender)
 
  
  
/* multiple regression */
regress write sex read math science
  
  
  
/* logistic regression */
logistic honor sex read math science
logit


/* factorial anova */
anova write gender prog gender*prog
  
    

How to cite this page

Report an error on this page