module metaprop for stata

[Habiba Bannour]

Hello,

Did anyone, please, use metaprop command within stata https://ideas.repec.org/c/boc/bocode/s457781.html. I m analysing different parameters of 2 groups of 15 RCTs. I could use for this the software called "comprehensive meta-analysis" but i want to calculate the risk ratios and the mean differences and i want draw the corresponding forrest plots with stata and metaprop, apparently, allows doing this. Can anyone explains me how we can do this meta-analysis Il's urgent!

Thank you very much, best regards,
Habiba BANNOUR

[Marcos Almeida]

Hello, "bannourh",

First things, first. Please re-register with name and family name (you can do this by clicking on the "contact us" button below to the right).

Second, I fear a forum for discussion on Stata doesn't necessarily apply to "urgent needs".

Third, since you (apparently) got the use-written metaprop from "search metaprop" in Stata, surely you also have full access to the help files as well (under "metaprop.sthlp").

Fourth, it seems you want the learn the whole process, I mean, you gave the impression you didn't just get stuck in a given point and want a tip to curb a precise problem. Therefore, I share with you an article from the Evidence -Based Mental Health (2014), aptly named "A hands-on practical tutorial on performing meta-analysis with Stata":

http://ebmh.bmj.com/content/17/4/111...mh_current_tab


I hope these suggestions help you just in time!

Best,

Marcos

[Habiba Bannour]

Thankyou.

1/ I re-registrated.

2/ I posted my question here because the facebook's admisnistrator proposed it to me (" You could also try posting your question on Statalist (http://www.statalist.org)") 3/ The help files, I read them yes but I didn't want to "program" the command, I wanted to use stata's buttons.

4/ Thank you for the article; I'm reading it.

I hope too, I have to finish my scientific poster for tomorrow. Thank you.

Habiba.

[Clyde Schechter]

First there is the question whether -metaprop- has a graphical user interface ("buttons"). Many user-written programs do not.

Assuming it does, using a graphical user interface to create a scientific poster or publication is not a very good idea, for several reasons. But given that you are under time pressure and don't feel you can learn to use a do-file quickly enough for this purpose, let me strongly encourage you to at least log all your work as you go. Otherwise, if you are at some point questioned about how you arrived at your results, you will have no record at all to refer to. And, weeks or months from now, you probably won't remember (and nobody will trust your memory even if you do.)

To open a log in the graphical user interface, click on the icon that looks like a tiny notebook (without a "pencil"). On my installation it appears just to the right of the printer icon. Follow the dialog boxes to create and save the log file under a convenient name in a convenient place. When you are done, clicking on that same icon will close the log. Make opening the log file the very first thing you do after launching Stata, in particular, before you open the data file. Then work until you have the results you need. Then finally close the log file and exit Stata. The log will contain a complete record of all the commands that Stata issued in response to your button-pushing, along with the resulting output.

[Marcos Almeida]

Dear Habiba,

Regarding your comment:

The help files, I read them yes but I didn't want to "program" the command, I wanted to use stata's buttons.

I wish to let you know that, unfortunately, the drop-down menu from Stata is not aimed at user-written commands.

Kind regards,

Marcos

[Nick Cox]

I'd say rather that on the whole user-programmers

1. do in the first instance write for themselves (and their students, etc.); they do not write to satisfy a market demand

2. tend to be highly command-oriented -- you can't program successfully otherwise -- and expect users to follow suit; alternatively, the view without malice or indifference is that others are more than welcome to write dialogs if they so wish

3. find it sufficiently challenging and amusing to produce fair quality code, test scripts and documentation, and so want to stop there. (A personal rule of thumb is that the help file requires about as much work as the code.)

This situation can be self-perpetuating. Without anyone saying so, among Stata user-programmers providing a program with a help file has widely become a self-evident minimal standard (exceptions don't proselytise), while conversely not providing a dialog is so common that few people even notice.

[Habiba Bannour]

Dear Clyde Schechter , I read this somewhere; I'm doing it (I "log" mes fichiers)
Dear Marcos Almeida , I'm understanding it, unfortunately.
Dear Nick Cox , if only stata could integrate this command in the official program (creating a graphical user interface for ut).
So, since the authors of this command didn't reyour advices lead me to return using the other meta-analysis' software: Comprehensive meta-analysis, isn't-it?
Thank you very much.
Habiba B.