Day 2-3: The
Difference Statistic and Modeling
You need to experiment to see if
ignoring two outcomes of a six sided die gives you the same results as
rolling a four sided die.
Roll a six sided die 50 times and
record how many 1s, 2s, 3s, and 4s you get. If you get 5 or a 6 do not
write anything down. After you have done this, copy the table below and
write the results in the appropriate places. Roll a 4 sided die the
same number as the total number of 1-4s you have. This will almost
definitely be less than 50. Record this information in the next column.
Are the results close to the same or they radically different?
Obtain the totals for the whole
class for six sided and four sided die rolls and see how those numbers
Why do you think the numbers should
be the same?
What are you, in effect, doing when
you ignore the five and six as outcomes on the six sided die?
How can you use this to do
preference testing with 5 products?
How could you use this with 3
products so that you don't need to ignore any of the outcomes?