{smcl}
{* 26Jun2006}{...}
{cmd:help multin}
{hline}
{title:Title}
{p 4 8 2}
{bf: multin -- Grouped conditional logit regression}
{title:Syntax}
{p 8 15 2}
{cmdab:multin:}
{depvar}
[{indepvars}]
{ifin}
{cmd:,}
{bind:{cmdab:gr:oup:(}{varname}{cmd:)} [{it:options}]}
{title:Description}
{p 4 4 2}
{cmd:multin} fits the grouped conditional logit regression model. It obtains the same results as conditional
logit regression but it requires a smaller data set. The data is prepared as for conditional logit regression but
each block of independent variables is unique and the dependent variable is a count with the number of times each
choice is selected. An example of a dataset prepared for estimation with {cmd:multin} is:
x1 x2 gid cid y
0 4 1 1 0
1 4 1 2 1
1 6 1 3 1
2 2 2 1 2
1 4 2 2 0
3 5 2 3 0
2 6 4 1 0
3 7 4 2 1
2 5 4 3 0
1 5 4 4 0
The variable gid is a unique group identifier and the dependent variable y shows the number of times that
each choice was selected. If your data is in the format required by clogit you can convert it to the above format
using {helpb groupdata}.
{title:Options}
{p 4 8 2}{cmd:addcon} Evaluates the likelihood with all constant factors added to the likelihood.
The results are the same but the reported likelihood is different.
{p 4 8 2}{cmd:pearson} reports the Pearson chi-squared statistic.
{title:Examples}
{p 4 8 2}{cmd:. multin y x1 x2, gr(id)}
{title:Author}
Paulo Guimaraes, Division of Research, University of South Carolina.
Email: {browse "mailto:guimaraes@moore.sc.edu":guimaraes@moore.sc.edu}
{title:Also See}
{p 4 13 2}
{helpb groupdata}
{helpb dirmul}