ttest.hlp

是一个经济学管理应用软件 很难找的 但是经济学学生又必须用到

{smcl}
{* 18mar2005}{...}
{cmd:help ttest, ttesti}{...}
{right:dialogs:  {bf:{dialog ttest1:one-sample}}  {bf:{dialog ttestpair:one-sample, paired}}}
{right:{bf:{dialog ttest2:two-sample}}  {bf:{dialog ttestby:two-sample, by()}}{space 2}}
{right:immediate dialogs:  {bf:{dialog ttesti1:one-sample}}  {bf:{dialog ttesti2:two-sample}}{space 8}}
{hline}

{title:Title}

{p2colset 5 18 20 2}{...}
{p2col:{hi:[R] ttest} {hline 2}}Mean comparison tests{p_end}
{p2colreset}{...}


{title:Syntax}

{pstd}
One-sample mean comparison test

{p 8 14 2}
{cmd:ttest}
{varname}
{cmd:==}
{it:#}
{ifin}
[{cmd:,} {opt l:evel(#)}]

{pstd}
Two-sample mean comparison test

{p 8 14 2}
{cmd:ttest}
{varname:1}
{cmd:==}
{varname:2}
{ifin}
[{cmd:,}
{it:{help ttest##options1:options1}}]

{pstd}
Group mean comparison test

{p 8 14 2}
{cmd:ttest}
{varname}
{ifin}
{cmd:,}
{opt by(groupvar)}
[{it:{help ttest##options2:options2}}]

{pstd}
Immediate form of one-sample mean comparison test

{p 8 14 2}
{cmd:ttesti}
{it:#obs}
{it:#mean}
{it:#sd}
{it:#val}
[{cmd:,}
{opt l:evel(#)}]

{pstd}
Immediate form of two-sample mean comparison test

{p 8 14 2}
{cmd:ttesti}
{it:#obs1}
{it:#mean1}
{it:#sd1}
{it:#obs2}
{it:#mean2}
{it:#sd2}
[{cmd:,}
{it:{help ttest##options3:options3}}]

{synoptset 16 tabbed}{...}
{marker options1}{...}
{synopthdr:options1}
{synoptline}
{syntab:Main}
{synopt:{opt unp:aired}}treat data as unpaired{p_end}
{synopt:{opt une:qual}}unpaired data have unequal variances{p_end}
{synopt:{opt w:elch}}use Welch's approximation{p_end}
{synopt:{opt l:evel(#)}}set confidence level; default is {cmd:level(95)}{p_end}
{synoptline}

{marker options2}{...}
{synopthdr:options2}
{synoptline}
{syntab:Main}
{p2coldent:* {opt by(groupvar)}}variable defining the groups{p_end}
{synopt:{opt une:qual}}unpaired data have unequal variances{p_end}
{synopt:{opt w:elch}}use Welch's approximation{p_end}
{synopt:{opt l:evel(#)}}set confidence level; default is {cmd:level(95)}{p_end}
{synoptline}
{p 4 6 2}* {opt by(groupvar)} is required.{p_end}

{marker options3}{...}
{synopthdr:options3}
{synoptline}
{syntab:Main}
{synopt:{opt une:qual}}unpaired data have unequal variances{p_end}
{synopt:{opt w:elch}}use Welch's approximation{p_end}
{synopt:{opt l:evel(#)}}set confidence level; default is {cmd:level(95)}{p_end}
{synoptline}
{p2colreset}{...}


{title:Description}

{pstd}
{opt ttest} performs t tests on the equality of means.  In the first form,
{opt ttest} tests that {varname} has a mean of {it:#}.
In the second form, {opt ttest} tests that {it:varname1} and {it:varname2}
have the same mean.  Data are assumed to be paired, but specifying
{opt unpaired} changes this assumption.  In the third form, {opt ttest} tests
that {it:varname} has the same mean within the two groups defined by
{it:groupvar}.

{pstd}
{opt ttesti} is the immediate form of {opt ttest}; see {help immed}.

{pstd}
For the equivalent of a two-sample t test with sampling weights
({opt pweight}s), use the {helpb "svy:mean"} command with the {opt by()}
option, and then use {helpb lincom}; also see {help svy postestimation}.

{title:Options}

{dlgtab:Main}

{phang}
{opt by(groupvar)} specifies the {it:groupvar} that defines the two groups
   that {opt ttest} will use to test the hypothesis that their means are
   equal.  Specifying {it:groupvar} implies an unpaired (two-sample) t test.
   Do not confuse the {opt by()} option with the {cmd:by} prefix; you can
   specify both.

{phang}
{opt unpaired} specifies that the data are to be treated as unpaired.  The
   {opt unpaired} option is used when the two set of values to be compared are
   in different variables.

{phang}
{opt unequal} specifies that the unpaired data not be assumed to have equal
   variances.

{phang}
{opt welch} specifies that the approximate degrees of freedom for the test
   be obtained from Welch's formula rather than Satterthwaite's approximation
   formula, which is the default when {opt unequal} is specified.  Specifying
   {opt welch} implies specifying {opt unequal}.

{phang}
{opt level(#)} specifies the confidence level, as a percentage, for confidence
   intervals.  The default is {cmd:level(95)} or as set by {helpb set level}.


{title:Examples}

{phang}{cmd:. ttest mpg == 20}{p_end}
{phang}{cmd:. ttest mpg1 == mpg2}{p_end}
{phang}{cmd:. ttest mpg1 == mpg2, unpaired}{p_end}
{phang}{cmd:. ttest mpg, by(treated)}{p_end}
{phang}{cmd:. ttest mpg, by(treated) unequal}

{phang}{cmd:. ttesti 24 62.6 15.8  75} (n=24,m=62.6,sd=15.8; test m=75){p_end}
{phang}{cmd:. ttesti 20 20 5  32 15 4} (n1=20,m1=20,sd1=5; n2=32,m2=15,sd2=4; test m1=m2)


{title:Also see}

{psee}
Manual:  {bf:[R] ttest}

{psee}
Online:  
{helpb bitest},
{helpb ci},
{helpb hotelling},
{helpb oneway},
{helpb prtest},
{helpb sdtest},
{helpb signrank},
{help svy postestimation},
{helpb "svy: mean"}
{p_end}