# Coefficient of variation question?

Is it appropriate to compare the coefficient of variation between discrete (e.g., clutch size) and continuous data (e.g., egg width)?

• Appropriate for what purpose? And how do you propose to make the comparison? – whuber Jul 17 '15 at 13:32
• A simple comparison to say, "clutch size was more variable than egg width in our population." Does this clarify my question? – user82580 Jul 17 '15 at 13:40
• Theoretically, a species should optimize their egg size and vary clutch size; therefore, there is a tradeoff between egg and clutch size. Some people essentially look at the coefficient of variation to determine if clutch size is more variable than egg size within a population to confirm the optimal egg size theory. My question is related to the appropriateness of making a comparison between a discrete and continuous variables'coefficient of variation. – user82580 Jul 17 '15 at 13:48

• I wonder about the equations you are making in this answer. The total energy expended by a bird to make a clutch would be pretty close to the total mass. In this sense, a difference of egg widths from $1.5$ to $2.5$ cm corresponds to a mass change of a factor $(2.5/1.5)^3 = 4.6$ whereas a change from $3.2$ to $4.2$ is a factor of $(4.2/3.2)^3 =2.3$: it's only half as big. This suggests abandoning CV as having any kind of meaningful relationship to the underlying biology and trying to exploit biological understanding to express the variables appropriately and make useful comparisons. – whuber Jul 17 '15 at 16:02