whatsnew8to9.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

{p 9 9 2}
     {cmd:vce(bootstrap)} specifies that the standard errors, significance
     tests, and confidence intervals be normal-based bootstrap estimates,
     rather than the default analytic estimates based on the observed
     information matrix.  You can also produce percentile-based or
     bias-corrected confidence intervals after estimation using 
     {cmd:estat bootstrap}; 
     see {helpb bootstrap postestimation:[R] bootstrap postestimation}.  

{p 9 9 2}
     {cmd:vce(jackknife)} specifies that the standard errors, significance
     tests, and confidence intervals be jackknife estimates.

{p 9 9 2}
     Both {cmd:vce(bootstrap)} and {cmd:vce(jackknife)} will automatically
     perform either observation or cluster sampling, whichever is appropriate
     for the estimator.

{p 9 9 2}
     Notably, both {cmd:vce(bootstrap)} and {cmd:vce(jackknife)} compute
     bootstrapped or jackknifed estimates of the complete VCE matrix.  This
     means that many of Stata's postestimation commands are available.  You
     can form linear and nonlinear combinations or functions of the parameters
     and obtain jackknife or normal-based bootstrap standard errors and
     confidence intervals for the combinations using {manhelp lincom R} and 
     {manhelp nlcom R}.  Similarly, you can perform linear and nonlinear tests
     using {manhelp test R} and {manhelp testnl R}.


{p 4 9 2}
12.  New command {cmd:estat}
     centralizes 
     the computing and reporting of additional statistics after 
     estimation, just as {cmd:predict} does with predictions.
     {cmd:estat} allows subcommands.  {cmd:estat} {cmd:summarize}, for
     instance, reports summary statistics for the estimation sample 
     and can be used after any estimator.
     {cmd:estat} also allows subcommands that are specific to the 
     estimation command.  To find out what is available after a command, 
     see the corresponding postestimation entry.  For example, after 
     {bf:[R] regress}, see 
     {help regress postestimation:{bf:[R] regress postestimation}};
     or after {bf:[XT] xtmixed}, see
     {help xtmixed postestimation:{bf:[XT] xtmixed postestimation}}.

{p 9 9 2}
    Existing postestimation commands have been brought into the
    {cmd:estat} framework:


	         Estimation       Old          New {cmd:estat}
	         command          command      command
	         {hline 50}
                 {cmd:regress}          {cmd:ovtest}       {cmd:estat} {cmd:ovtest}
                                  {cmd:hettest}      {cmd:estat} {cmd:hettest}
                                  {cmd:szroeter}     {cmd:estat} {cmd:szroeter}
                                  {cmd:vif}          {cmd:estat} {cmd:vif}
                                  {cmd:imtest}       {cmd:estat} {cmd:imtest}
        
                 {cmd:regress}          {cmd:dwstat}       {cmd:estat} {cmd:dwatson}
                 (time series)    {cmd:durbina}      {cmd:estat} {cmd:durbinalt}
                                  {cmd:bgodfrey}     {cmd:estat} {cmd:bgodfrey}
                                  {cmd:archlm}       {cmd:estat} {cmd:archlm}
        
                 {cmd:anova}            {cmd:ovtest}       {cmd:estat} {cmd:ovtest}
                                  {cmd:hettest}      {cmd:estat} {cmd:hettest}
        
                 {cmd:logit} and        {cmd:lstat}        {cmd:estat} {cmd:classification}(*)
                 {cmd:logistic}         {cmd:lfit}         {cmd:estat} {cmd:gof}(*)
        
                 {cmd:poisson}          {cmd:poisgof}      {cmd:estat} {cmd:gof} 
        
                 {cmd:stcox}            {cmd:stphtest}     {cmd:estat} {cmd:phtest}

                 {cmd:xtgee}            {cmd:xtcorr}       {cmd:estat} {cmd:wcorrelation}
	         {hline 50}
		 (*) The new command works after {cmd:probit}, as well
                     as {cmd:logit} and {cmd:logistic}; the old command worked
                     after {cmd:logit} and {cmd:logistic} only.

{p 9 9 2}
The original commands continue to work but are undocumented.

{p 9 9 2} 
    Three {cmd:estat} subcommands are available after almost all estimators:

{p 9 13 2}
    a.  {cmd:estat} {cmd:ic} reports Akaike's and Schwarz's Bayesian
        information criteria (AIC and BIC).

{p 9 13 2}
    b.  {cmd:estat} {cmd:summarize} reports summary statistics on the
        variables in the estimation model for the estimation sample.

{p 9 13 2}
    c.  {cmd:estat} {cmd:vce} reports the covariance (VCE) or correlation
        matrix estimates.  ({cmd:estat} {cmd:vce} replaces the old 
        {cmd:vce} command and has more features.)

{p 4 9 2}
13.  Stata has many new prefix commands (commands that behave like {cmd:by:}
     and {cmd:xi:}).  New prefix commands include {cmd:statsby:},
     {cmd:bootstrap:}, {cmd:jackknife:}, {cmd:permute:}, {cmd:simulate:},
     {cmd:stepwise:}, {cmd:svy:}, and {cmd:rolling:}.  For instance, to obtain
     the standard error and confidence interval of the mean, you might type

		{cmd:. jackknife: mean earnings}

{p 9 9 2}
     or to obtain survey-adjusted estimates, you might type 

		{cmd:. svy:  mean earnings}

{p 9 9 2}
     after {cmd:svyset}ting your data.

{p 9 9 2}
    See 
    {help bootstrap:{bf:[R] bootstrap}},
    {help jackknife:{bf:[R] jackknife}},
    {help permute:{bf:[R] permute}},
    {help rolling:{bf:[TS] rolling}},
    {help simulate:{bf:[R] simulate}},
    {help stepwise:{bf:[R] stepwise}},
    {help statsby:{bf:[D] statsby}}, 
    and
    {help svy:{bf:[SVY] svy}}.

{p 4 9 2}
14.  New prefix commands {cmd:bootstrap:} and {cmd:jackknife:} replace old
     commands {cmd:bs} and {cmd:jknife}, and in addition to having better
     syntax, they also provide new features:

{p 9 13 2}
    a.  They handle and report of expressions better.

{p 9 13 2}
    b.  They post their results as
        estimation results with a complete VCE.  Most postestimation
        facilities may now be used after them and will be based on 
        the bootstrap or jackknife VCE.  These include

{p2colset 17 28 30 0}{...}
{p2col:{helpb adjust}}adjusted predictions{p_end}
{p2col:{helpb estimates}}cataloging estimation results{p_end}
{p2col:{helpb lincom}}linear combinations with SEs, tests, and CIs{p_end}
{p2col:{helpb nlcom}}nonlinear combinations with SEs, tests, and CIs{p_end}
{p2col:{helpb mfx}}computing marginal effects and elasticities{p_end}
{p2col:{helpb predict}}predictions, residuals, probabilities, etc.{p_end}
{p2col:{helpb predictnl}}generalized nonlinear predictions with SEs and CIs{p_end}
{p2col:{helpb test}}Wald tests of simple and composite linear hypotheses{p_end}
{p2col:{helpb testnl}}Wald tests of nonlinear hypotheses{p_end}
{p2colreset}{...}

{p 9 13 2}
    c.  They produce a model test when
        applied to the coefficients of estimation commands.

{p 9 13 2}
    d.  They allow option {opt seed(#)} to set the random number seed.

{p 9 13 2}
    e.  They allow option {opt reject(exp)} to reject replicates that
        explicitly match {it:exp}.

{p 9 13 2}
    f.  {cmd:bootstrap:} uses the normal distribution
        instead of the Student's t distribution to compute the
	normal-approximation confidence intervals.

{p 9 13 2}
    g.  {cmd:jackknife:} now allows {cmd:fweight}s to be specified.

{p 9 9 2}
See
    {help bootstrap:{bf:[R] bootstrap}} and
    {help jackknife:{bf:[R] jackknife}}.


{p 4 9 2}
15.  New prefix command {cmd:statsby:} replaces old command {cmd:statsby} (not
     a prefix) and provides enhanced handling and reporting of expressions,
     allows {cmd:weights}, and allows string variables in the option 
     {opt by()}.
     See {help statsby:{bf:[D] statsby}}.

{p 4 9 2}
 16.  New prefix command {cmd:stepwise:} replaces old command 
      {cmd:sw} and, in addition to working with all the previous estimators,
      also works with {helpb intreg:[R] intreg} and {helpb scobit:[R] scobit}.

{p 4 9 2}
17.  Existing prefix command {cmd:xi:} has new option {opt noomit} that
     prevents it from omitting a category when generating category indicators
     for group variables.
     See {helpb xi:[R] xi}.

{p 4 9 2}
18.  New command {cmd:tetrachoric} computes a tetrachoric correlation
    matrix for a set of binary variables.  See
    {helpb tetrachoric:[R] tetrachoric}.

{p 4 9 2}
19.  Existing command {cmd:suest}, which combines estimation results for
    subsequent testing, is easier to use and has new features:
	
{p 9 13 2}
        a.  Scores are now computed for the models you
            combine; you no longer need to save scores when
	    estimating.

{p 9 13 2}
        b.  {cmd:suest}, used after {cmd:svy:} estimation, now accounts 
           for your survey design.

{p 9 13 2}
        c.  {cmd:suest} now works more smoothly with certain estimation
            commands that previously required special treatment, including 
            {helpb regress}, {helpb ologit}, and {helpb oprobit}.

{p 9 13 2}
        d.  {cmd:suest} now works with all models estimated by {cmd:clogit},
	    rather than only those with a single positive outcome per group.

{p 9 9 2}
        See {help suest:{bf:[R] suest}}.

{p 4 9 2}
20.  Existing command {cmd:clogit} has new features:

{p 9 13 2}
	a.  Robust and cluster-robust SE and VCE estimates are now supported
	    through options {cmd:robust} and {cmd:cluster()}.

{p 9 13 2}
	b.  Linear constraints on the parameters are now implemented via
	    option {cmd:constraints()}.

{p 9 13 2}
	c.  New option {cmd:vce()} allows SE and VCE estimates to be
	    computed using OIM (the default), OPG, bootstrap, and jackknife.

{p 9 9 2}
    See {help clogit:{bf:[R] clogit}}.

{p 4 9 2} 
21.  Option {cmd:level()} now allows noninteger
     confidence levels to be specified.
     See {help estimation options:{bf:[R]} estimation options}.

{p 4 9 2}
22.  Existing command {cmd:predict} now generates equation-level scores
     after most maximum-likelihood estimation commands; see the documentation
     of {cmd:predict} in the postestimation entry for each estimation command.

{p 4 9 2}
23.  Existing command {help cumul} has a new option {cmd:equal} to create
     equal cumulative values for ties.  See {help cumul:{bf:[R] cumul}}.

{p 4 9 2}
24.  Existing command {cmd:estimates table} now allows you to specify more
     models, and the command wraps the table if necessary.  
     Also allowed are new options

{p 9 13 2}
        a.  {cmd:equations()}, which matches equations by number rather than
            by name.

{p 9 13 2}
        b.  {cmd:coded}, which displays the table in a compact, symbolic
            format.

{p 9 13 2}
        c.  {cmd:modelwidth()}, which sets the number of characters for
            displaying model names.

{p 9 9 2}
See {helpb estimates:[R] estimates}.

{p 4 9 2}
25.  {cmd:test} after {helpb anova} and {helpb manova} has two new options
        for performing Wald tests:

{p 9 13 2}
    a.  {cmd:mtest()}, which implements three methods to adjust for multiple
        tests: Bonferroni, Holm, and Sidak.

{p 9 13 2}
    b.  {cmd:test()}, which makes specifying contrasts easier by accepting a
         matrix containing the contrast.

{p 9 9 2}
See {helpb anova postestimation:[R] anova postestimation}.

{p 4 9 2}
26.  Commands {cmd:ci} and {cmd:cii} have new options {cmd:exact},
     {cmd:wilson}, {cmd:agresti}, {cmd:jeffreys}, and {cmd:wald} for computing
     different types of binomial confidence intervals.
     See {help ci:{bf:[R] ci}}.

{p 4 9 2}
27.  Command {cmd:hausman} has new option {cmd:df()} for controlling the
     degrees of freedom.  See {help hausman:{bf:[R] hausman}}.

{p 4 9 2}
28.  {cmd:predict} after {helpb ivreg} has the new {opt score} option for
     returning equation-level scores.
     See {help ivreg postestimation:{bf:[R] ivreg postestimation}}.

{p 4 9 2}
29.  Command {cmd:mfx} is now faster and has new option {cmd:varlist()} for
     computing effects of specific variables.  See {help mfx:{bf:[R] mfx}}.

{p 4 9 2}
30.  Commands {cmd:tabulate} and {cmd:tabi} with the {cmd:exact} option are
     now significantly faster.

{p 4 9 2} 
31.  In existing command {cmd:mlogit}, option {cmd:basecat} has been renamed
     {cmd:baseoutcome()} for better consistency with the terminology of choice
     models.
     See {helpb mlogit:[R] mlogit}.

{p 4 9 2} 
32.  Existing commands {cmd:spearman} and {cmd:ktau} now allow more than
    two variables to be specified and have more flexible output.
    See {helpb spearman:{bf:[R] spearman}}.

{p 4 9 2}
33.  Existing command {cmd:bsample} for sampling with replacement
    (bootstrap sampling