inasero
Reborn
- Joined
- Nov 27, 2002
- Messages
- 2,497
- Gender
- Male
- HSC
- 2003
yes i agree wholeheartedly! when did i suggest otherwise? you do recognise i was being sarcastic...
hey who's a n00b?! methinks u are t3h n00bi3...think about it- there are differences in cohorts from year to year, even within the same university using the same selection criteria...look at my year and the year above for example...
my second point (supporting the fact that u r a n00b) rests on the fact that you hung yourself with your own rope
. If we can't split a cohort in half cos "the specimens aren't supposed to know which group they are", then how on earth can we use two different year levels studying different methods? That's in essence committing the same experimental errors except with twice the number of people. Thirdly, your proposal that we could compare two year levels is shot down by my splendiferous observation being this- you're saying that my model is "disastrous" because the two groups would know who they were (i.e. you can't teach a blind curriculum to a cohort of students). But then how about your model? Say students never knew what a PBL/PCL was and were blinded...the experimenters would certainly know and would bring bias into the evaluation...ergo my orginial proposal that we can't have a "control" group as such even when odds adjustment and confidence intervals are taken into account, you said it yourself- "(not good because of too many uncontrolled environmental variables)."
muaAHahAHhAH feel the wrath of my epidemiology!!!!!!!!
p|^|nz01231)
Last edited:
