Cumulative Relative Frequency vs Relative Frequency

Cumulative relative frequency tells you what fraction of the data you have accounted for by the time you reach a certain point, so it rises to 1 as you work through the ordered categories. You compute it as $\frac{\text{cumulative count}}{n}$, where $n$ is the total number of observations. In a class of 40 students, if 10 scored below 70 and another 14 scored in the 70s, the cumulative count through the 70s is 24 and the cumulative relative frequency is $24/40 = 0.60$. That 0.60 is also a statement about position: the top of the 70s band sits at roughly the 60th percentile.

Relative frequency rescales a raw count into a fraction or percent, which makes groups of different sizes comparable. For example, if 18 of 40 students walk to school, the relative frequency is $18 / 40 = 0.45$, or 45 percent. The relative frequencies across all categories add up to $1$, that is, 100 percent. This idea is the empirical basis for estimating probabilities from data.