{smcl}
{* 19Aug2008}{...}
{hline}
help for {hi:metaeff}
{hline}

{title:Title}

{p2colset 5 16 18 2}{...}
{p2col :{hi:metaeff} {hline 2}}Meta-analysis module for effect sizes calculations{p_end}
{p2colreset}{...}


{title:Syntax}

{p 4 8 2}
{cmd:metaeff}
{it:varname1}
{it:varname2}
{ifin}
[{cmd:,} {it:{help collapse##options:options}}]

{p 4 4 2}
where

{p 6 6 2}
{it:varvame1} the variable corresponding to the estimated effect sizes.

{p 6 6 2}
{it:varvame2} the variable corresponding to the estimated standard errors of the effect sizes.

{synoptset 20 tabbed}{...}
{marker options}{...}
{synopthdr}
{synoptline}
{syntab :Options}
{synopt :{opth ni(varname)}}specifies the number of units in the intervention group
{p_end}
{synopt :{opth nc(varname)}}specifies the number of units in the control group
{p_end}
{synopt :{opth i(varname)}}specifies the number of events in the intervention group
{p_end}
{synopt :{opth c(varname)}}specifies the number of events in the control} group
{p_end}
{synopt :{opth orval(varname)}}specifies the odds ratio
{p_end}
{synopt :{opth ci95lo(varname)}}specifies the lower end of 95% confidence interval of the mean difference or odds ratio
{p_end}
{synopt :{opth ci95up(varname)}}specifies the upper end of 95% confidence interval of the mean difference or odds ratio
{p_end}
{synopt :{opth md(varname)}}specifies the mean difference (not standardised)
{p_end}
{synopt :{opth meani(varname)}}specifies the mean of the intervention group
{p_end}
{synopt :{opth meanc(varname)}}specifies the mean of the control group
{p_end}
{synopt :{opth sdi(varname)}}specifies the standard deviation in the intervention group
{p_end}
{synopt :{opth sdc(varname)}}specifies the standard deviation in the control group
{p_end}
{synopt :{opth ci95loi(varname)}}specifies the lower end of the 95% confidence interval of the mean for the intervention group
{p_end}
{synopt :{opth ci95upi(varname)}}specifies the upper end of the 95% confidence interval of the mean for the intervention group
{p_end}
{synopt :{opth ci95loc(varname)}}specifies the lower end of the 95% confidence interval of the mean for the control group
{p_end}
{synopt :{opth ci95upc(varname)}}specifies the upper end of the 95% confidence interval of the mean for the control group
{p_end}
{synopt :{opth p(varname)}}specifies the p value for a two-sample t-test
{p_end}
{synopt :{opth t(varname)}}specifies the t value for a two-sample t-test (assumes df=ni+nc-2)
{p_end}
{synopt :{opt or}}selects odds ratio (1b) over the default risk difference method (1a), when ni, nc, i & c are available
{p_end}
{synopt :{opt owrite}}specifies that {it:varvame1} and {it:varvame2} will be overwritten, if they already exist in the dataset
{p_end}
{synopt :{opt infovar}}provides information on the method selected (for each case) in a new string variable _method
{p_end}


{title:Description}

{p 4 4 2}
The {cmd:metaeff} command provides a way to calculate the effect sizes (and the respective standard errors)
of research studies, for use with meta-analysis methods (see {cmd:{help meta}}, {cmd:{help metan}} & {cmd:{help metaan}}).
The methods used for the calculations have been derived from the Cochrane Collaboration handbook. The calculated effect sizes are
standardised mean differences and binary data are transformed towards that end. {cmd:metaeff} is meant to be used for meta-analysing studies
with various types of outcomes; if all studies provide binary outcomes then {cmd:metaeff} should not be used since better approaches exist and a
transformation is not required.

{p 4 4 2}
{hi:Dichotomous data}

{p 6 4 2}
{hi:Method 1a}, based on Risk Difference. Requires {hi:ni}, {hi:nc}, {hi:i} & {hi:a}.

{p 6 4 2}
{hi:Method 1b}, based on Odds Ratio. Requires {hi:ni}, {hi:nc}, {hi:i} & {hi:a}.

{p 6 4 2}
{hi:Method 2}, based on Odds Ratio. Requires {hi:ni}, {hi:nc}, {hi:orval}, {hi:ci95lo} & {hi:ci95up}.

{p 4 4 2}
{hi:Continuous data}

{p 6 4 2}
{hi:Method 3}, based on Mean Difference. Requires {hi:ni}, {hi:nc}, {hi:md} or ({hi:meani} and {hi:meanc}) , {hi:ci95lo} & {hi:ci95up}.

{p 6 4 2}
{hi:Method 4}, based on Mean Difference. Requires {hi:ni}, {hi:nc}, {hi:md} or ({hi:meani} and {hi:meanc}), {hi:sdi} & {hi:sdc}.

{p 6 4 2}
{hi:Method 5}, based on Mean Difference. Requires {hi:ni}, {hi:nc}, {hi:md} or ({hi:meani} and {hi:meanc}), {hi:ci95loi}, {hi:ci95upi}, {hi:ci95loc}, {hi:ci95upc}.

{p 6 4 2}
{hi:Method 6}, based on Mean Difference. Requires {hi:ni}, {hi:nc}, {hi:md} or ({hi:meani} and {hi:meanc}) & {hi:p} or {hi:t}.

{p 4 4 2}
{hi:Dichotomous or Continuous data}

{p 6 4 2}
{hi:Method 7}, based on z-value. Requires {hi:ni}, {hi:nc} & {hi:p} or {hi:t}.


{title:Remarks}

{p 4 4 2}
Some methods are more accurate than others and have been prioritised accorgingly (in case user has provided input that allows
2 or more methods to be selected). For dichotomous data the method order is {hi:2}, {hi:1a}, {hi:1b}, {hi:7} and for continuous
data {hi:4}, {hi:5}, {hi:3}, {hi:6} & finally {hi:7}. Methods {hi:6} and (especially) {hi:7} should only be used if no other
method can be selected (due to distibutional assumptions and the use of the reported p-value is usually inaccurate). The methods are described
in Kontopantelis and Reeves, 2009 and have been prioritised according to expected precision: that is, the expectation that the effect size and
associated variance computed from the input data will be accurate. As a general rule, the fewer the number of mathematical transformations involved
in getting from the 'raw data' to the statistical parameters used as input for the method, the higher the expected precision.

{title:Examples}

{p 4 8 2}
{cmd:. metaeff eff SEeff, ni(ivsize) nc(clsize) i(ivevents) c(ivevents)}

{p 4 8 2}
{cmd:. metaeff eff SEeff, ni(ivsize) nc(clsize) i(ivevents) c(ivevents) or owrite}

{p 4 8 2}
{cmd:. metaeff eff2 SEeff2, ni(ivsize) nc(clsize) md(meandiff) sdi(ivsd) sdc(clsd)}

{p 4 8 2}
{cmd:. metaeff eff SE, ni(ivsize) nc(clsize) meani(ivmean) meanc(clmean) ci95loi(ivlo) ci95upi(ivup) ci95loc(cllo) ci95upc(clup)}

{p 4 8 2}
{cmd:. metaeff eff SEeff, ni(ivsize) nc(clsize) t(tvalue)}


{title:Authors}

{p 4 4 2}
Evangelos Kontopantelis, National Primary Care Research and Development Centre,

{p 29 4 2}
University of Manchester, e.kontopantelis@manchester.ac.uk

{p 4 4 2}
David Reeves, Health Sciences Primary Care Research Group, University of Manchester


{title:References}

{p 4 4 2}
Cochrane Collaboration handbook: {browse "http://www.cochrane.org/resources/handbook":http://www.cochrane.org/resources/handbook/}
Kontopantelis, E. and Reeves D. 2009.  A Meta-Analysis add-in for Microsoft Excel.  Journal of Statistical Software.


{title:Also see}

{p 4 4 2}
STB: STB-44 sbe24

{p 4 4 2}
help for {help metaan}, {help metan7} (if installed)

{p 4 4 2}
{help metannt} (if installed), {help meta} (if installed)

{p 4 4 2}
{help metacum} (if installed), {help metareg} (if installed)

{p 4 4 2}
{help metabias} (if installed), {help metatrim} (if installed)

{p 4 4 2}
{help metainf} (if installed), {help galbr} (if installed)

{p 4 4 2}
{help metafunnel} (if installed)