This package contains 3 functions used for implementing the Mean Score method
of Reilly & Pepe (Biometrika 1995, see full reference below) to
estimate the coefficients in a logistic regression model from two-stage data.

   1. MEANSCORE is called with the combined first- and second-stage data
      (where the missing covariate values are represented by NA)
   2. MS.NPREV is called with the second-stage (i.e. complete) data 
      and the first-stage sample sizes (or prevalences). Prior to running
      this function, the CODING function (3.) should be run to see the
      the order in which MS.NPREV expects the first-stage sample sizes
      or prevalences to be provided.
   3. CODING, which recodes multiple columns of first-stage covariates into 
      a single column and displays the coding scheme

Two illustrative data sets are also provided

  simNA   a simulated data set of 1000 observations, with 500 missing
          values (represented by NA)
 
  ectopic the data used in Table 3 of Reilly and Pepe (1995) 


Help on these functions and on the illustrative data sets
provided can be viewed using help or ? or the HTML help file system.


Reilly,M and Pepe MS 1995. A mean score method for missing and auxiliary 
      covariate data in regression models. Biometrika 82:299-314
