Statistics Help @ Talk Stats Forum

regression and log poisson

Delphine Van Ruysten — Thu, 23 May 2013 11:34:38 GMT

Hi,

This is my first post here.

I would like to know if this code:

GLM.1 <- glm(y ~ sex + sex * I(age-39), family=poisson(log), data=Dataset)

is corresponding to this model:

http://imageshack.us/a/img267/4431/c...0515154134.png

(I have to intepret thoses parameters)

Cause i got some results but i'm not sure it's the right code for the model. If it's wrong, wich program do you suggest me?

Here we have the 5 first lines of my dataset:

y sex age
4 0 50
0 0 21
6 0 43
0 0 30
0 0 26

Thanks for answers!

D.V.R.

Normal curve and averages

sfischer — Thu, 23 May 2013 10:52:35 GMT

Hi,

I am hoping this is a simple one. I have a normally distributed population with mean of 20 and standard deviation of 10. I want to look only at the portion of the population between 15 and 30. I then want to work out (or approximate at least) the mean of this portion of the population. I can work out what percentage of the upper and lower tail that was cut off but I am unsure what to do with that information to recalculate my new average. Can someone please help - thanks.

Statictics Assistance needed

Mischa23 — Thu, 23 May 2013 09:38:23 GMT

Hi Im really stuck with a particular question regarding calculating confidence intervals. I dont expect anyone to solve it for me but just a push in the right direction would help heaps :)

Recent research published in the Proceedings of the National Academy of Sciences has
examined whether judges make different decisions regarding whether or not to grant
parole just after a break compared to a long time after a break. Suppose the
researchers gathered data on 150 rulings from first thing in the morning and 110
rulings from just before the end of the day. You can assume random ordering of
parole cases throughout the day (in other words, parole cases for more serious
offences are equally likely to appear in the first thing in the morning or late in the
day). Of the early morning cases 96 were granted parole, of the cases considered just
before the end of the day 22 were granted parole.

Calculate a two-sided 95% confidence interval for the difference between the
proportions of those grated parole first thing in the morning (F) and those granted
parole late in the day (L). Interpret the results.

I think its, (0.44) +- 1.96 x the square root of the Std Deviation/260. Is that right?

Any help would be massively appreciated!