Question 1:

Consider the software training problem. That is, consider the 
problem where a training consultant disagrees with the widely held
notion that training casual users of basic software tools like
Wordprocessors and Spreadsheets is ineffective. She believes 
that a short training session can dramatically improve the 
efficiency of casual Microsoft Word users and decides to 
conduct an experiment to investigate this issue.

She randomly selects two groups of casual Microsoft Word users 
from her organization and decides to use one group as a Control 
group and the other group as a Treatment group. That is, the 
Control group will not receive training but each member in the 
Treatment group will receive a short training session on various 
Microsoft Word features. 

She assigns a suite of tasks to each individual in each group 
and records the time required to complete the suite.  

Let us say that the following statistics are available:
   Control Group: n=49; ybar=56; s(y)=18
   Treatment Group: n=64; ybar=50; s(y)=20
Given these details, answer the following: 

a) Provide a point estimate for mu(t)-mu(c)
b) Construct and interpret an 80% confidence interval for your
   estimate.
c) Construct and interpret a 95% confidence interval for your
   estimate.
d) Let us say that n(c)=20 instead of 49 and n(t)=16 instead of
   64. Conduct a test of hypotheses.
Note: Make and state any necessary assumptions.



Question 2:

Consider the software inspections problem. That is, let us say
you are interested in comparing two different techniques (say 
method a and method b) that may be used to conduct software 
inspections. Many of your colleagues argue that the techniques
are equally effective but, based on personal experience, you 
contend that method b is better than method a and decide to 
design an experiment to settle the matter. 

To evaluate the methodologies you decide to randomly select 
two groups of programmers from your software development 
team. You train each group on one of the techniques and then 
give them a few days to get comfortable with the technique. 
Each programmer is then given a suite of modules to inspect 
and the number of software bugs detected is recorded. 

a) Given the problem statement above, identify and state the 
   null and alternative hypotheses. 
b) Let us say that the following statistics are available:
   method a: n=20; ybar=18; s(y)=6 
   method b: n=20; ybar=23; s(y)=4
   Given these details and your hypotheses, conduct a test of 
   hypotheses. Remember to comment on the respective techniques.
c) Given your answer to b), do you think a point estimate of
   the difference in poulation means is appropriate? If so,
   construct and interpret a 90% confidence interval. 
Note: Make and state any necessary assumptions.