o talking about blocking
  block= techniqued used to reduce variability
  by grou;ing alike exp. units
ex: beads & tod
- so far
 block-treatment model
Y(hit)=m+O(h)+t(i)+e(hit)
  where t=1 since randomized
complete block
- now general complete block
t is not 1 since s>1
as above with no interaction
 - have interaction?
y(hit)=m+O(h)+t(i)+(Ot)(hi)+e(hit)
h=1,...b
i=1....v
t=1,...s
- these look like and
 analyze like 2-way main effects
 and 2-way complete
- proper questions
1. H0: no interaction
2. H0: blocks do not help
3. H0: no treatment effect
note: if interaction, do 
cell-means follow-ups

- note: check assumptions
- conduct appropriate follow-ups
- ex:
 bloc-treatment interaction model
see printout
  * have interaction so
do cell-means follow-up...see
waht erroneously testing 
H0: no size effect 
does

1	1	60    Source               SS          DF        MS       F     p-val
1	1	55    SIZE     		18.7        2           9.3     .84   .478                  
1	2	39    TOD      		85.3        1          85.3   7.64   .033
1	2	42    SIZE*TOD        1034.7        2         517.3   46.33  .001       
1	3	30    Error             67.0        6         11.17       
1	3	35    Total           1205.7       11       
2	1	35
2.	1.	40    means size 1 = 47.5 tod 1=43.5  sizextod 11=57.5     21=37.5
2.	2.	50               2=46.5       2=48.3         12=40.5       22=52.5
2.	2	55               3=44.5                      13=32.5       23=56.5
2.	3.	59
2.	3.	54