lol in my data the strongest negative correlation i found with "how many sexual partners have u had" is "what % of the porn you watch is animated?" at r=-0.17
Nonetheless, the pool she is pulling from is far too narrow to be statistically significant for the broad population. With that said, it's still big enough to make you wonder (it's not surprising from a gut reaction standpoint).
Statistical significance ≠ validity, and besides i am pretty sure this example has external validity as well.
there are religious celibate people who don't watch porn at all and are unlikely to have participated, but one could still check the effect in anime porn [...]