[R-meta] About meta-regression analysis

[Viechtbauer, Wolfgang (NP)]

Dear Nick,

One does not have to manually change a categorical moderator into a numerical one to examine interactions. Of course, categorical moderators are represented in a model in terms of numerical variables (typically via a set of dummy variables), but R can handle all of this encoding for you automatically. Here is an example with two categorical variables, one consisting of two levels, the other with three levels:

https://www.metafor-project.org/doku.php/tips:multiple_factors_interactions

Best,
Wolfgang

> -----Original Message-----
> From: R-sig-meta-analysis <r-sig-meta-analysis-bounces using r-project.org> On Behalf
> Of ?????? via R-sig-meta-analysis
> Sent: Monday, December 11, 2023 08:38
> To: r-sig-meta-analysis using r-project.org
> Cc: 英文科陳品誠 <t571 using wlgsh.tp.edu.tw>
> Subject: [R-meta] About meta-regression analysis
>
> Hello all,
>      As I read through the "Doing meta in R" website and to the part of
> meta-regression, this is the example that the website provides in order to
> change the categorical moderators into numerical ones (since you need
> "numbers" to begin to test the interaction between moderators):
> --------------------------------------------------------------------
> # Add factor labels to 'continent'
> # 0 = Europe
> # 1 = North America
> levels(MVRegressionData$continent) = c("Europe", "North America")
> --------------------------------------------------------------------
> and I thought of my own data. I have a moderator in my data named "Duration
> of treatment," with "Short", "Medium", "Long" and "N/A"(some of the studies
> didn't provide the duration of treatment). In this case, how would I code
> in order to change this moderator into numerical order to begin the
> meta-regression analysis?
> Nick
> 陳品誠 (Nick Chen)
> Email: t571 using wlgsh.tp.edu.tw <t5741 using wlgsh.tp.edu.tw>