DD had just turned 7 when she took the CogAT in second grade, and her age percentiles were only slightly higher than grade. For instance for verbal she may have been at 97th percentile for age but 95th percentile for grade. She took a version of the test that is timed and normally given to third graders, so I'm not sure if that made any difference. It is optional for schools to give kids a higher-level version of the test so that the test ceiling is higher. Her grade scores were compared to other second graders though, not third graders.
DS took the CogAT and the grade and age percentiles were identical. He has a late winter birthday so he is probably middle of the pack in terms of age for his grade. When he took it, he was 7 years, 8 months, I think. It seems a little weird that you are seeing such big differences in terms of the percentiles when I saw very little difference with DD, who was very young compared to most other second graders who were tested (she was grade-accelerated). You might want to try to verify that they didn't make an error with her birthday or grade.
