I'm doing a study about the correlation between working memory and math grades for students. And I have to do a one-way ANOVA test (or some other ANOVA?).

I already have the math grades, but the working memory score I have to calculate from a working memory test. The test is performed by having the participants doing the same test with the same test conditions 4 times.

My question is: what's the best/correct way to calculate the working memory scores for the ANOVA test? As I see it there are 2 choices:

1) Take the average of the 4 tests, for every participants and use that value for my working memory score in my ANOVA test.

2) Regard all four tests are done by 4 different participants. Then I use all the values instead of the average.

Or should I have a completely different approach such as repeated measures ANOVA, ANOVA with replication etc...?