@karlahemming If small number of pairs, suspect Hartung-Knapp would be appropriate: https://t.co/4RJI3Cs6AN
@giovanni_ef @TobiasS25347008 @BelavySpine @zadro_josh @PhysioMeScience @JOSPT @PatrickOwenPhD @HugoPedder @msk_health Number of studies (k) is used in the calculation of HKSJ CIs. You can view the steps on page 6 in this example: https://t.co/vyjYNMNEOG
@giovanni_ef @PhysioMeScience @JOSPT @PatrickOwenPhD @BelavySpine Dear Mr. Ferreira, i would recommend the following sources for the justification of taking 5 trials. They also give an indepth overview of the topic. https://t.co/xYyRbgS8ax https://t.co/
RT @Richard_D_Riley: @_MiguelHernan @xabieradrian @oriolmitja For random-effects meta-analysis, Hartung-Knapp Sidik-Jonkman derived 95% con…
RT @Richard_D_Riley: @_MiguelHernan @xabieradrian @oriolmitja For random-effects meta-analysis, Hartung-Knapp Sidik-Jonkman derived 95% con…
RT @Richard_D_Riley: @_MiguelHernan @xabieradrian @oriolmitja For random-effects meta-analysis, Hartung-Knapp Sidik-Jonkman derived 95% con…
RT @Richard_D_Riley: @_MiguelHernan @xabieradrian @oriolmitja For random-effects meta-analysis, Hartung-Knapp Sidik-Jonkman derived 95% con…
@_MiguelHernan @xabieradrian @oriolmitja For random-effects meta-analysis, Hartung-Knapp Sidik-Jonkman derived 95% confidence intervals have better coverage (as account for uncertainty in variance estimates). Worth adding to the results (will lead to much
As such it generally leads to much better coverage (https://t.co/Xc8mU39fDe), and so I recommend it, though is not a perfect solution (https://t.co/9tZh8APCf5 https://t.co/9X0XSuL0Y6)
A better alternative is the Hartung-Knapp-Sidik-Jonkman method. It controls the false positive rate more effectively, at the cost of fewer "statistically significant" findings! https://t.co/vvtUVuhdLK
RT @aroradrn: @icmjc @f_g_zampieri I found this... and it made my headache worse... https://t.co/pizigIziov #ICMJC
@icmjc @f_g_zampieri I found this... and it made my headache worse... https://t.co/pizigIziov #ICMJC
@AgniBankai Did you mean: https://t.co/RFDq7y6gT2
Hartung-Knapp-Sidik-Jonkman method for random effects meta-analysis outperforms the standard DerSimonian-Laird method http://t.co/NSCebUycFj