tips:comp_two_independent_estimates

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

Both sides previous revision Previous revision | |||

tips:comp_two_independent_estimates [2019/05/22 08:42] Wolfgang Viechtbauer |
tips:comp_two_independent_estimates [2019/05/22 08:54] (current) Wolfgang Viechtbauer |
||
---|---|---|---|

Line 1: | Line 1: | ||

===== Comparing Estimates of Independent Meta-Analyses or Subgroups ===== | ===== Comparing Estimates of Independent Meta-Analyses or Subgroups ===== | ||

- | Suppose we have summary estimates (e.g., estimated average effects) obtained from two independent meta-analyses or subgroups of studies and we want to test whether the estimates are different from each other. A Wald-type test can be used for this purpose. Alternatively, one could run a single meta-regression model including all studies and using a dichotomous moderator to distinguish the two sets. Both approaches are conceptually very similar with a subtle difference that will be illustrated below with an example. | + | Suppose we have summary estimates (e.g., estimated average effects) obtained from two independent meta-analyses or two subgroups of studies within the same meta-analysis and we want to test whether the estimates are different from each other. A Wald-type test can be used for this purpose. Alternatively, one could run a single meta-regression model including all studies and using a dichotomous moderator to distinguish the two sets. Both approaches are conceptually very similar with a subtle difference that will be illustrated below with an example. |

==== Data Preparation ==== | ==== Data Preparation ==== | ||

Line 49: | Line 49: | ||

</code> | </code> | ||

- | We can now compare the two estimates (i.e., the estimated average log risk ratios) by feeding them back to the ''rma()'' function and using the variable to distinguish the two estimates as a moderator. We use a fixed-effects model, because the (residual) heterogeneity within each subset has already been accounted for by fitting random-effects model above. | + | We can now compare the two estimates (i.e., the estimated average log risk ratios) by feeding them back to the ''rma()'' function and using the variable to distinguish the two estimates as a moderator. We use a fixed-effects model, because the (residual) heterogeneity within each subset has already been accounted for by fitting random-effects models above. |

<code rsplus> | <code rsplus> | ||

rma(estimate, sei=stderror, mods = ~ meta, method="FE", data=dat.comp, digits=3) | rma(estimate, sei=stderror, mods = ~ meta, method="FE", data=dat.comp, digits=3) | ||

Line 81: | Line 81: | ||

-1.395 | -1.395 | ||

</code> | </code> | ||

- | This is the same value as obtained above. | + | This is the same value that we obtained above. |

==== Meta-Regression with All Studies ==== | ==== Meta-Regression with All Studies ==== | ||

- | Now let's take a different approach, fitting a meta-regression model using all studies: | + | Now let's take a different approach, fitting a meta-regression model with ''alloc'' as a categorical moderator based on all studies: |

<code rsplus> | <code rsplus> | ||

rma(yi, vi, mods = ~ alloc, data=dat, digits=3) | rma(yi, vi, mods = ~ alloc, data=dat, digits=3) |

tips/comp_two_independent_estimates.txt ยท Last modified: 2019/05/22 08:54 by Wolfgang Viechtbauer

Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-Noncommercial-Share Alike 4.0 International