A marginal relative frequency in a two-way frequency table is the ratio of a marginal frequency and the total number of values or observations. As an example, the following data will be used.
Hat preference | ||||
Top hat | Beret | Total | ||
Gender | Male | $6$ | $12$ | $18$ |
Female | $15$ | $20$ | $35$ | |
Total | $21$ | $32$ | $53$ |
Dividing the frequencies by $53$ will give the joint and marginal relative frequencies.
Hat preference | ||||
Top hat | Beret | Total | ||
Gender | Male | $536 ≈0.11$ | $5312 ≈0.23$ | $5318 ≈0.34$ |
Female | $5315 ≈0.28$ | $5320 ≈0.38$ | $5335 ≈0.66$ | |
Total | $5321 ≈0.40$ | $5332 ≈0.60$ | $1$ |
The blue entries in the table are the marginal relative frequencies. They show, for instance, that about $66%$ of all participants of the survey are female. Notice that, ignoring the error margin introduced by rounding, a marginal relative frequency can be found by adding a row or column of joint marginal frequencies.