Graphical interface for estimating effects
effmenu1 [varlist] [if exp] [in range] [, clear display ]
Description
Command with menus for evaluating the effect of exposure on a response, possibly modified by a modifying variable, and possibly controlled for one or more controlling variables. When a varlist is specified, only those variables in the varlist are included in the menus.
The command produces a menu in which to specify
The response variable often known as the outcome The type of response binary/metric/failure/count The follow-up time for survival data The matched set variable for matched case-control studie > s The exposure variable i.e. the main explanatory variable The modifying explanatory variable if required Whether study is matched case-control for matched case-control studies Matched set indentifier for matched case-control studies Baseline baseline level for a categorica > l exposure
When either the exposure or modifier is categorical, a maximum of 15 values is > allowed.
Selecting Effects produces a sub-menu in which to specify how the effects are to be measured according to the response:
metric differences of means/ratios of means binary ratios or differences of proportions/odds ratios failure rate ratios/rate differences count rate ratios/rate differences
WARNING: when selecting ratios of proportions, differences in proportions, or rate differences, there may be problems in convergence with some data sets.
This sub-menu gives the opportunity to change the units of the effect to per # when the exposure variable is metric. The sub-menu also allows the inclusion of control variables, which can be categorical or metric. If they are metric they must be selected in the both the control variables box and in the control variables which are metric box.
Options
clear Clears previous selections from the menu.
display Produces effmenu2, the command line version of effmenu1.
Also see
Help for effmenu2, tabmenu1, tabmenu2.