- Because we can’t sample every part.
- Because no matter what we do, no two parts will be the same.
- Because we define a product to meet specifications and we can measure how well it conforms.
- Because differences between parts are hard (assignable causes) or impossible (chance causes) to predict.
- Because we can sample a few and draw conclusions about the whole group.
- We want to sample as little data as possible to draw the most accurate conclusions about the distribution of the values.
- Inductive - Try to get overall variance within group. i.e. assume all of group should conform ***WE USE THIS TYPE
- Deductive - Attempt to classify differences that exist within a group (i.e. election polls).