I am looking for others' thoughts and experiences regarding the variability on the AMC8 tests. AMC8 testing is coming up next week. Since DS/DD do not belong to any type of math club and I wanted their exposure to tricky math problems as well as have some preparation for the test, they have taken a half dozen of the past AMC8 tests over the last month and a half. Interestingly, the past stats (average, top 5%, top 1%, etc) on the AMC8 are very consistent from year to year. However, my children's scores are not. For example, on a good day, DS may make the top 1% of all scorers, but on a bad day, he may barely make the category for high scoring 6th (and below) graders. It is clear that some of it has to do with relative carelessness (based on redo of incorrect problems) and some of it has to do with the particular distribution of types of problems (DS has more trouble with geometry), but there appears to be some variability beyond that.

Has anyone else observe that type of variability - a difference of 6 problems (out of 25)?