Review sheet for final exam april 22, 2002 5:30 p.m.

 remember to come with homework problems chapter 6 #7 & #8, c. 7 #5, 
ch. 10 #6,
 ch 11 #11, ch. 13 #2 #8 (omit b), ch. 15 #1

 o be able to plot interaction including those for 3-way and interpret
 o be able to read a SAS printout (3 or higher factorial) and correctly
 interpret when
there are equal and unequal sample sizes (LSMEANS and TYPE III SS)
 o be able to interpret SAS multiple comparisons such as scheffe, tukey
  and bonferroni and dunnett as well as construct these follow-ups by 
hand (given the MSE)
 o understand the use of table A2 to build orthogonal contrasts and how 
to do and use
 a plot like that on p. 215 figure 7.7
 o be able to apply the "taguchi approach" to an analysis
 o from SAS output be able to decide whether blocking is helpful in 
an experiment
 and whether the 
 treatments differ and also be able to interpret multiple comparisons 
via scheffe,
 tukey, bonferroni and  dunnet; be able to do follow-ups by hand
 (given MSE)
 o in a balanced incomplete block design be able to explain why the LS 
estimates of
 mean + theta + tau is not Y-bar (hi.).  
be able to generate one tau and one theta by hand
 from a small table of observations.
 o be able to say which levels of the experiment (0000 etc) 
must be conducted in a
 2**p factorial design
 blocked into 2 or 2**2 blocks in order to confound a specific effect.
 o be able to work with SAS output for a confounded 2-level 
factorial experiment
 o be able to use the technique of confounding to develop a
 2-level fractional factorial
 design and  identify which effects are aliased; and to use the 
contrasts to generate effect
 estimates by hand in a small experiment