Before you put this lesson away, let's look back at what you have
learned about taste tests and how to do them.
Points to remember:
1. Before doing a taste test you must assume that all products are
equally liked by the general public. If you do not assume this, it is
called a bias, and this will change the outcome of your test. On your
own, think about how this could be true.
2. After you obtain your data, the difference which you need
to begin your quest to finding the d-statistic, is the difference
between the votes received by a particular product and the expected
number of votes the product should have received if all were preferred
equally. This number is always (Total number of votes) divided by
(Total number of products).
3. Follow the sample tables and their order of operations to
get the d-statistic. Remember, you are not finished here. Without
comparing it to random d-statistics which fit the mold of your taste
test you can not declare a winner.
4. You need to generate several d-statistics randomly using
die, coins, a random number table, etc. to see how large your
d-statistic is compared to ones produced by chance. Remember, you need
as many outcomes as you had possible products. If you had seven
products, you cannot use a die because it can only have six outcomes.
Get it? If not, ask for clarification from a teacher or peers. If yours
happens to be larger than 90% or more than the randomly produced ones
then it is said to be significant and you may declare a winner.