ááááá name:á <unnamed>

áááááá log:á C:\Documents and Settings\Michael Rosenfeld\My Documents\newer web pages\soc_meth_proj3\2010_logs\third_class_log.log

á log type:á text

áopened on:áá 2 Feb 2010, 14:55:44


. use "C:\Documents and Settings\Michael Rosenfeld\Desktop\cps_mar_2000_new.dta", clear


. table yrsed if age>24 & age<35, contents(freq mean yrsed sd yrsed)

rowvar variable(s) may not be used in contents()



. table sex if age>24 & age<35, contents(freq mean yrsed sd yrsed)



ááááá Sex |áááááá Freq.á mean(yrsed)ááá sd(yrsed)


áááá Male |áááááá 9,027ááá á13.31212áááá 2.967666

áá Female |áááááá 9,511áááá 13.55657áááá 2.854472



. display 13.55657-13.31212



. *the difference between men and women's education in this age group is about one quarter of a year. Is that a big difference? Are we sure the women have more education?



. ttest yrsed if age>24 & age<35, by(sex)


Two-sample t test with equal variances


áá Group |áááá Obsá ááááááMeanááá Std. Err.áá Std. Dev.áá [95% Conf. Interval]


ááá Male |ááá 9027ááá 13.31212ááá .0312351ááá 2.967666ááá 13.25089ááá 13.37335

á Female |ááá 9511ááá 13.55657ááá .0292693ááá 2.854472ááá 13.49919ááá 13.61394


combined |áá 18538ááá 13.43753ááá .0213921ááá 2.912627áááá 13.3956ááá 13.47946


ááá diff |áááááááááá -.2444469ááá .0427623áááááááááááááá -.3282649áá -.1606289


ááá diff = mean(Male) - mean(Female)ááááááááááááááááááááááááááááá t =á -5.7164

Ho: diff = 0áááááááááááááááááááááááááááááááááááá degrees of freedom =ááá 18536


ááá Ha: diff < 0áááááááááááááááá Ha: diff != 0áááááááááááááááá Ha: diff > 0

áPr(T < t) = 0.0000áááááááá Pr(|T| > |t|) = 0.0000ááááááááá Pr(T > t) = 1.0000


* The t-statistic of -5.7, being much larger in absolute value than 2, demonstrates that this difference of .244 years between men and women in years of education is powerfully significantů See my Excel file where this problem is worked out by hand.



. exit, clear