{smcl}
{* 20dec2007}{...}
{cmd:help blogit_2P}
{hline}

{title:Title}

{p2colset 5 15 21 2}{...}
{p2col :{cmd:blogit_2P} {hline 1}} Logistic regression for grouped two-phase data {p_end}
{p2colreset}{...}


{title:Syntax}

{phang}
Logistic regression for grouped two-phase data

{p 8 16 2}{cmd:blogit_2P}{space 2}{it:dep_var} {it:stratum_var} {it:P1count_var} {it:P2count_var}  [{it:rhsvars}] 
   [{cmd:,} {it:{help glogit_2P##options:options}}]

{phang}

{synoptset 20 tabbed}{...}
{marker options}{...}
{synopthdr :options}
{synoptline}

{syntab : Analysis type}
{synopt :{opt method(analtype)}}{it:analtype} may be {opt WL} or by default {opt ML} {p_end}

{syntab :Reporting}
{synopt :{opt design}}reports the detailed two-phase design {p_end}

{synoptline}


{title:Description}

{pstd}
{cmd:blogit_2P}  produces maximum-likelihood logit(ML) or weighted likelihood (WL)
estimates on grouped ("blocked") two-phase data.  

In the syntax diagrams, {it:dep_var} is a binary indicator variable  of a positive response and 
{it:stratum_var} is an integer variable containing a stratum number varying from 1 to the total
number of strata.
  
{it:P1count_var} contains the number of subjects in Phase 1 corresponding to the given combination
of the dependent variable and the stratum variable (NB : this number is identical for each 
combination of independent variables within the stratum).

{it:P2count_var} contains the Phase 2 number of subjects corresponding to the given combination of
independent variables within the given combination of the dependent variable and the stratum 
variable 



{title:Examples}

{pstd}Logistic regression for grouped two-phase data fitted by ML using the EM algorithm with 
independent variable {it:exposure}{p_end} {phang2}{cmd:. blogit_2P  disease z_stratum Nij nijk exposure }{p_end}

{pstd}Same as above, but fitted by WL {p_end}
{phang2}{cmd:. blogit_2P  disease z_stratum Nij nijk exposure,method(WL) }{p_end}


{pstd} Another logistic regression with independent variables {it:expo1 expo2 expo3} fitted by WL, 
but displaying the complete two-phase design {p_end}
{phang2}{cmd:. blogit_2P  disease z_stratum Nij nijk expo1 expo2 expo3,method(WL) design}{p_end}

{title:Saved results}

{pstd}
{cmd:blogit_2P} saves the following in {cmd:e()}:

{synoptset 20 tabbed}{...}
{p2col 5 20 24 2: Scalars}{p_end}
{synopt:{cmd:e(Nstrata)}}number of strata{p_end}
{synopt:{cmd:e(Ncases1)}}number of cases ({it:dep_var=1}) in Phase 1{p_end}
{synopt:{cmd:e(Ncontrols1)}}number of controls ({it:dep_var=0}) in Phase 1{p_end}
{synopt:{cmd:e(Ncases2)}}number of cases ({it:dep_var=1}) in Phase 2{p_end}
{synopt:{cmd:e(Ncontrols2)}}number of controls ({it:dep_var=0})  in Phase 2{p_end}

{synoptset 20 tabbed}{...}
{p2col 5 20 24 2: Macros}{p_end}
{synopt:{cmd:e(method)}}{cmd:ML} or {cmd:WL}{p_end}
{synopt:{cmd:e(depvar)}}name of dependent variable i.e. the disease indicator{p_end}
{synopt:{cmd:e(strata)}}name of variable containing the strata number{p_end}
{synopt:{cmd:e(P1counts)}}name of variable containing the Phase One counts{p_end}
{synopt:{cmd:e(P2counts)}}name of variable containing the Phase Two counts{p_end}


{synoptset 20 tabbed}{...}
{p2col 5 20 24 2: Matrices}{p_end}
{synopt:{cmd:e(b)}}coefficient vector{p_end}
{synopt:{cmd:e(V)}}variance-covariance matrix of the estimators{p_end}