sdm-help-list > Jackknife analysis on meta-regression
Showing 1-2 of 2 posts
Jun 4, 2015 02:06 PM | Nobody
Jackknife analysis on meta-regression
Hi all,
Just a quick one. Is is possible to carry out a jackknife/sensitivity analysis on the results from a meta-regression/multiple meta-regression in SDM? A number of meta-analytical papers appear to allude to this but it is not clear how it was carried out (if at all).
Many thanks
Jack
Just a quick one. Is is possible to carry out a jackknife/sensitivity analysis on the results from a meta-regression/multiple meta-regression in SDM? A number of meta-analytical papers appear to allude to this but it is not clear how it was carried out (if at all).
Many thanks
Jack
Jun 22, 2015 12:06 PM | Nobody
RE: Jackknife analysis on meta-regression
Dear Jack,
easy jackknife on SDM is only implemented for the mean analysis. You can however carry out jackknife for any analysis using filter variables, e.g. creating as many filter variables as studies, setting these filters to be 1 for all studies except for one, and repeating the analyses applying these filters. You can also set this to be automatic using scripts that modify the sdm_table.txt file and call SDM.
Best wishes,
Joaquim
easy jackknife on SDM is only implemented for the mean analysis. You can however carry out jackknife for any analysis using filter variables, e.g. creating as many filter variables as studies, setting these filters to be 1 for all studies except for one, and repeating the analyses applying these filters. You can also set this to be automatic using scripts that modify the sdm_table.txt file and call SDM.
Best wishes,
Joaquim
