Bella (Isabella) Smith
Biostatistics/Bayesian Statistics
HW 1
1. Probabilities from the slides about lung cancer
Below is the MATLAB code we ran:
a.Sally: hard evidence in the code above
tl=30, k=114; p=0.2632
b.John: simply changed hard evidence and last
Since our slope B1 is positive, this means that e^Bi is also positive and greater than one. This means that
as Ni increases, the logit(pi) increases, which means that pi increases.
Since our slope B1 is positive, this means that e^Bi is also positive and
Bella Smith
3) Model is in WINBUGS attachment.
a) the Bayes estimators of the 5 priors are listed above under the mean column.
b) the 95% credible set for the Jeffreys prior is given by the 2.5% and 97.5% stats. The 95%
credible set for the Jeffreys prior
1. A/B)
Above the one way ANOVA was run for the difference in measurements for the mirrors.
To determine the treatment levels that are different, in the table we look for the
intervals that contain zero. If the interval contains 0, then we say they are no