Indeed you can! And Lady Larunai did not.Fallow said:Of course population sizes are comparable, that's why we do the whole proportion and scaling thing.
Hatecrime is a category of crimes, like violent crime or financial crime. You can scale that to populations by taking number of victims as a proportion of the available victim population and compare it to another category by doing the same there.
If you need more detail, you can scale the populations by age-to-victim probabilities and so on.
You can do all kinds of fun stuff with statistics.
You've approached the population issue, which was only relevant insofar as it invalidated Lady Larunai's proposed percentage. You've not addressed the issue of the numbers being unrelated.Fallow said:Of course, that relies on hatecrime being bounded by availability of victims, which seems overly pessimistic. If we don't separate the category from the others we make the assumption that hatecrime is not limited by availability of perpetrators and victims, nor that we have to specify perp-to-victim relations. That is also an imperfect solution, but far from a gamebreaker. Thus, using proportions make perfect sense.
Once again: if we go by this metric, then hate crime becomes less severe as we criminalise more things. Murder may be deemed insignificant as it represents a small proportion of all crimes; crime may be deemed insignificant as it represents a tiny proportion of all events.
This is-- yet again-- restating a weakness of the evidence. Fine. It is not indicating that MarsAtlas was using the absence of numbers as evidence, which was your suggestion. That is still pure misrepresentation.Fallow said:And that is exactly it. We don't know the extent. If we don't know the extent, how can you claim it's significant? How can you say it's a factor when you don't know? And, since hatecrime does not exist in a vacuum, how can you claim that it's a significant factor in relation to the underreporting of every other crime? If all crime is underreported by 30% then your factor is 1.0.