Use data set 'three'.

We want to predict the amount of radiation detected

on a mouse, based on the strain of the mouse (A or B) and the number of hours

of exposure.

A dummy variable, d, has been created for the strain, where

strain B is the referent level.

i) Run a regression model that predicts radiation from the dummy variable and

hours.

Obtain a plot of residuals vs. hours;

turn in

this plot and comment on

what you see.

ii) Using the model in part (i), use the CLM option to obtain a 95% confidence

interval for the mean radiation for strain A when hours = 1, and again when

hours = 9.

Report these values.

Question: How to account for both strain and hours, a sas code please

iii) The natural log of radiation is the variable 'lograd'.

Using this

transformation of the dependent variable, run a regression model that predicts

'lograd' from the dummy variable and hours.

Obtain a plot of residuals vs.

hours;

turn in

this plot and comment on what you see.

iv) Using the model in part (iii), again use the CLM option to obtain a 95%

confidence interval for the mean radiation for strain A when hours = 1, and

again when hours = 9.

Note that SAS will give you the predicted value and CI

endpoints for the

log

of radiation.

You will have to "back transform" this to

radiation by exponentiating these values.

Show how you did this!

Please provide the sas code