Hi StataList,
I've done multivariate meta-analysis:
Is it possible (or make sense) to have study weights, forest plot, funnel plot, leave-one-out diagnostics after multivariate meta-analysis (not separately for A_lin and A_exp but jointly) ? if yes how?
Regards
I've done multivariate meta-analysis:
Code:
meta mvregress A_lin A_exp , wcovvariables(A_lin_var A_lin_exp_cov A_exp_var )
Code:
* Example generated by -dataex-. For more info, type help dataex
clear
input str25 study float(A_lin A_exp A_lin_var A_lin_exp_cov A_exp_var)
"{b1}" 1.0285043 -.5900092 1.443407 2.544972 60.37708
"{b2}" .02 1.3634e-09 .000684633 -.04205276 14.121278
"{b3}" .8470645 -.09021278 .6838545 .3518806 52.7144
"{b4}" .14061001 -.02191767 1.221233 -1.564513 60.91925
"{b5}" .08758754 .03399021 1.1623507 -1.2919837 61.62936
"{b6}" .96 -.08 1.6927164 -.01117672 .008434296
"{b7}" .115427 -.05093065 85031.52 14.564894 62.8355
"{b8}" .6722152 -.20525235 2.7307556 .001159497 .08728556
"{b9}" .6688729 -.2046574 3.567039 -.000409764 .10452537
"{b10}" -.008867991 .11533568 .3307522 -.005029278 .009878436
"{b11}" .3791344 -.1743777 5141463 1044.6276 669.1354
"{b12}" -.8342307 -1.7570702 1432662 3481.0574 2038934.3
"{b13}" .034285873 .2511763 .42368445 .0000664143 .161404
"{b14}" .3434255 .05056002 .26622882 -.03853713 .016389543
"{b15}" 31.22262 -.28610772 145932.95 -11.834876 .6135564
"{b16}" 16.471739 -.3989164 4393.838 -23.215744 .3078139
end
Is it possible (or make sense) to have study weights, forest plot, funnel plot, leave-one-out diagnostics after multivariate meta-analysis (not separately for A_lin and A_exp but jointly) ? if yes how?
Regards
