Can "effect size" be determined without also knowing the population standard deviation?

What type of statistics are you using, parametric or nonparametric? The former requires a measure of variation like the standard deviation.