RM ANOVA on a frequency distribution?

Hi -

I fell over a research project in which frequency distributions from 3 different groups were analysed by use of repeated measures ANOVA.

The data were divided into a number of bins (similar to a histogram), and the dependent variable was the number of observations in each bin as a percentage.
The repeated independent variable was the bins and the (non-repeated) independent variable was group. Each individual would, obviously, sum up to 100%.

In addition, the cumulative frequency was analysed in a similar way with the cumulative frequency as the dependent variable. Again, the repeated factor was the bin and the non-repeated factor was the group.

To me, these approaches appear to be wrong. At least the latter approach, since each bin will include all of the data that preceded it. Consequently, data from the first bins are weighted more heavily than those from the later bins and will contribute disproportionately to the significance of the difference between the groups.
But is the first approach acceptable?

Thank you.