We'll know that our sample average is not the same as the real average, there’s no easy way to know when our guess is too high or too low.
This variation between the sample average and the overall average we’ll call bias.īecause of within-group variation and bias, comparisons among groups become harder. Another important point is that we won’t expect the average strength of our sample to be the same as the average strength if we taped a million boxes. The differences in strength measurements from the same supplier’s tape give us within-group variation. Instead, we’ll measure the strength from a sample of taped boxes and use those measurements to guess what the numbers would look like if we taped a million boxes.Īn important point is that we won’t expect all the measurements in a group to be the same.Ĭonsider the tape example again. But if we taped those million boxes and measured the peel strength, we would have used up all of the tape. If you could tape 1 million boxes from a batch of tape, those million might represent the entire population that we want to know about.
Inferential analysis is the formal way of saying that we want to look at a sample of measurements and make an educated guess about what all of the possible measurements might be like if we could take them. All the strength measurements for the same supplier’s tape form a group of measurements.ĪNOVA is an inferential statistical analysis. For example, if you want to know whether tapes from three different suppliers have the same peel strength, the suppliers are your factor. While ANOVA has many varieties, the essential purpose of this family of analyses is to determine whether factors have an association with an outcome variable.įactors are the variables that you will use to categorize your outcome variable into groups.
0 kommentar(er)
