edit: to be clear, I am not saying that there is without a doubt no correlation between self-identified furries and zoophilic desires or that the rate of those with zoophilic desires within a solely self-identified furry population is not higher than the base rate of zoophilic-urges in non self-identified furries, I am saying that within this population sampled the papers findings are accurate, relevant and true.
The fact is that that data just doesn't exist in the fashion discussed here, with a nuanced breakdown of the sub-paraphilias and the SIA-RA indicators correlations, for the 'regular' non furry, non zoophile population.
One doesn't need the data, the sentence appears to assert a one way correlation and that is not coherent.
If you have populations (A) 10,000, (B) 1000, (C) 100. A contains B and C. C and B may overlap or they may not. Obviously B is 10 times larger than C.
If B and C don't overlap that is perfect negative correlation, being in B means you are not C, being C means you are not B.
C cannot contain B but B can contain C. If B contains C then being C implies you are B (100% chance), but there is only a 10% chance that a randomly selected B is C. This is maximum positive correlation between different sized groups.
In this case being B entails a significantly higher likelihood of being C than the likelihood of being C in the general population (A). To put numbers to that a randomly selected B has a 10% chance of being C, but a randomly selected A has a 1% chance of being C.
One last point of interest is zero correlation. Zero correlation is when knowing that someone is B does not help you predict whether they are C or not. That occurs when the overlap between B and C has the exact same ratio to B as the overlap between C and A. Obviously C is contained in A so the ratio is 100:10,000 = 1:100, which is just a restatement of the 1% chance of A being C.
The zero correlation overlap between B and C is thus 10:1000 = 1:100,
that is only 10 people are both C and B. There is a 1% chance that a B is a C and a 1% chance that an A is a C. What about from C to B? 10/100 Cs are Bs so there is a 10% chance that a random C is a B, and in the general population there is a 10% chance of being a B.
All other possibilities exist on a continuum between these three scenarios.
Now let's put some words to the letters.
A = general population
B = furies
C = zoos
The quote was:
That is, zoophilia indicates a higher likelihood of furryism, however, furryism does not necessarily entail a greater likelihood of zoophilia.
Translated to variables:
That is, C indicates a higher likelihood of B, however, B does not necessarily entail a greater likelihood of C.
Note that "greater likelihood" is in reference to the likelihood of being zoo or furry in general i.e. for the general population.
If "C indicates a higher likelihood of B" then C is positively correlated with B, which is to say the overlap between B and C relative to A is greater than the ratio between B and A.
Let the likelihood of being a furry = X. Let the likelihood of being a zoo = Y.
The likelihood of a furry being zoo = Xy. The likelihood of a zoo being a furry = Yx.
Zero correlation between furies and zoos implies: Xy = Y and Yx = X.
Negative correlation between furies and zoos implies Xy < Y and Yx < X.
Positive correlation between furies and zoos implies Xy > Y and Yx > X.
The claim of the quoted statement is: Yx > X, but Xy <= Y. This cannot be as illustrated above. You can have positive correlation or negative correlation or no correlation but you can't have more than one at the same time.
I don't know if anyone has done this poll but it seems like maybe half of zoos would self-ID as furry. I would, with the caveat of complaining that the definition is very loose. Half the general population is not furry. That's positive correlation. Now if the furry group is significantly larger than the zoo group we could be talking insignificant changes in likelihood, but it probably isn't. Let me throw some ballpark figures for example:
Yx = 0.5 [what I just asserted]
Y = 0.005 [Kinsey study said 2% I think probably an overestimate; 0.5% is a nice conservative number]
Now we just need and X, not so easy I found some numbers after a search but they're probably wrong like ~60,000 total. The prevalence of zoosexuality is of course global being an artifact of humanity living in a world with non-humans which is a lot more universal than people brought up on animated anthro animal movies who have access to the internet and can go to conventions in the western world. If you start to count anthro gods, legends, or talking animals in oral tradition you have an argument for a more universal definition of "furry".
So let's pretend that we're only talking about 1st world countries with the appropriate generation for maximum furiness and assert X = 8%. I think it would be fairish to say that 8% of say a modern college campus has some furiness going on.
So:
Y = 0.005
X = 0.08
Yx = 0.5
Xy = Yx * total zoos / total furies = Yx * Y*A / X * A = Yx * Y / X = 0.5 * 0.005/0.08 = 0.03125
Which is to say a furry would have a 3.1% chance of being a zoo while the general public has a 0.5% chance of being a zoo.
Or in other words furies are 6x more likely to be zoos than the general population.