Differences

This shows you the differences between two versions of the page.

--- tips:convergence_problems_rma [2019/09/28 10:26] – external edit 127.0.0.1
+++ tips:convergence_problems_rma [2022/03/26 14:50] (current) – Wolfgang Viechtbauer
@@ Line 179: / Line 179: @@
 </code>
+Such differences are (usually) negligible and not practically relevant.
 ==== Maximum Likelihood Estimation ====
-The examples above all involved the use of restricted maximum-likelihood (REML) estimation, which is the default for models fitted with the ''rma()'' and ''rma.mv()'' functions. Regular maximum likelihood (ML) estimation can be used with ''method="ML"''. Let's try this out with the data from the previous example.
+The examples above all involved the use of restricted maximum likelihood (REML) estimation, which is the default for models fitted with the ''rma()'' and ''rma.mv()'' functions. Regular maximum likelihood (ML) estimation can be used with ''method="ML"''. Let's try this out with the data from the previous example.
 <code rsplus>
@@ Line 215: / Line 216: @@
 </code>
-Same value up, at least up to 5 decimals places. In fact, if we increase the accuracy (i.e., lower the threshold) and round even less, we still get the same values with the EB and PM estimators.
+Same value, at least up to 5 decimals places. In fact, if we increase the accuracy (i.e., lower the threshold) and round even less, we still get the same values with the EB and PM estimators.
 <code rsplus>