--start---

biostat again. oh argh argh argh.
if we stay on schedule our last homework will be distributed THIS 
THURSDAY.

the remaining material can be characterized in two capsules. some of the 
capsules are pretty cursory. you need to see it, try to understand it, 
but not be "responsible" for it. eg, today's material (which might flow 
into wednesday) shoudl be seen and understood, but is mroe complex (HA!) 
so you don't really need to be able to do the nuts and bolts how does it 
work stuff. so, for handout six, enjoy it but don't go nuts with it. the 
regression stuff is also like that. but the material in handout 7 on 
categorical data (chi squared test, etc) should be paid MORE attention, 
because we'll need it again in epidemiology.

note also that handout five had a lot of material in it from an 
arithmetic and philosophical standpoint. we are now going to embellish on 
that, applying it to different venues in the examination of data. the 
models will be a bit different but basically stemming from the same 
premise.


in the t test for two groups we compared data under varying situations.

Ho: mu1 = mu2 was the null for both naturally occuring and controlled 
groups.

when groups were controlled, mu were not naturally occuring attributes, 
but were numbers deemed to be the average response under exhaustive 
testing conditions.

what happens if there are more than two groups you need to compare? maybe 
there are several herds of cows or you're comparing several doses of a 
drug or several versions of a procedure to each other or to a control.

control == placebo == something that looks like drug but has no drug in 
it

positive control == a procedure or drug that is already being use, 
against which your test thing is being tested. this is osmething that 
actually works and is more commonly used these days.

Ho: mu1 = mu2 = ... = muk
Hi: all mu are not equal

Ho: all mu equal
Hi: all mu NOT equal

you gather data in following manner
previously you only had two groups, here there are k groups. for each 
group you will gather observations as before. but now we need subscripts. 
oh, btw, this is in handout 6. 

oh this is awful.
I'm so lost. I need to stop taking notes, and just read the handout for a 
while.

		()___()
        / o o \
        \     /___________====\\
        /     \           \  
        \_0_0_/            \
           \               |
            |  |--------|  |
            |  |        |  |    
            ----        ----
---end----