Open christinabergmann opened 5 years ago
This can be done straight from metafor but it would mean that we would have to change over from plotly. It would be less interactive. It would look like the funnel plots from MAJOR.
We can also just fix the current plot, no need to rip out the current plotting code.
On Wed, Jun 12, 2019 at 12:14 PM W. Kyle Hamilton notifications@github.com wrote:
This can be done straight from metafor but it would mean that we would have to change over from plotly. It would be less interactive. It would look like the funnel plots from MAJOR.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/langcog/metalab2/issues/134?email_source=notifications&email_token=AAI25F2CWKP2OHPAZOMXHKLP2DEAJA5CNFSM4HNEF2XKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODXP5WMY#issuecomment-501209907, or mute the thread https://github.com/notifications/unsubscribe-auth/AAI25F35IFFB5T3QJGFGHPLP2DEAJANCNFSM4HNEF2XA .
That'd be ideal to fix the current plot. We basically need to re-center the triangles to 0 and could add a line for the estimated meta-analytic mean. (I also realized the description is misleading, waiting to adapt though).
I think it would be better to show funnel plots such that the shaded area indicates non/significance (we could have a dotted line for the current funnel around the estimates mean effect) - i.e. that the center is around 0.
Details here: http://www.metafor-project.org/doku.php/plots:contour_enhanced_funnel_plot?s[]=contour&s[]=enhanced