Posted by Llurpsie_Noodle on February 27, 2007, at 11:43:40
In reply to Impressed and bedazzled!!!!! » Llurpsie_Noodle, posted by zazenduckie on February 26, 2007, at 22:48:40
> What is "improved outcome" and how are you measuring it?
That's the million dollar question. My personal favorite measure is improvement on the babbleometer scales. Dr. Bob was smart to select Babbleometer scales like the CES-D that are widely reported in the literature, and that people in the mental health field know well.
>
> I like that regression thing too like we could ask everyone who they were in past lives and see if they sort themselves out by historical era like do all the Cleopatras and Egyptians mostly post to each other and are all the admin regulars 18th century rationalists or the psych board 19th century romantics etc etc etcMy pick for 19th century romantic is probably Honore or Racer- although Honore is also dangerously close to being part of Dada.
Personally I'd like to regress to the womb. I believe my mom took me to France when I was prenatal. That sounds pretty good, huh?
> I'm glad you're feeling sparklier :)
You noticed! well, I'm glad too. really glad. starting to feel more llurpsie and less limp noodle.
happy day to you zazenduckie :)
-Ll
>
>
> > I think we should use heirarchical linear modelling to examine whether posters who identify with one particular "board" on babble show improved outcome relative to other posters.
> >
> > poster nested within board repeated measures, of course. that will let us look at outcomes for people who participate on different boards. Are there group differences, or do individuals account for most of the variability in outcome.
> >
> > can you tell I've been stuck in statistics land too much?
> >
> > (((((((data)))))))
> >
> > but Dr. bob, before analyses can be run, we gotta get data.
> >
> > welcome back, by the way. I missed you. and please change the picture. it makes me feel dizzy?
> >
> > what are the Dependent Variables? what are the Independent variables. I LOVE me some statistics. Data don't lie. people do. If dependent variables are not quantified yet, then they will need to be coded. I think that this would be a real boon to the literature. I don't know how to write up a case report (unless it's my own, in which case I'd say psychobabble is a good thing)
> >
> > now I'm just procrastinating
> >
> > ((((((student SPSS)))))))
> >
> > I just splurged and got an upgrade to do advanced regression, mixed and linear models. now I'm super-woman. don't mess with me!
> >
> > -Ll
>
>
poster:Llurpsie_Noodle
thread:709392
URL: http://www.dr-bob.org/babble/admin/20070123/msgs/736780.html