3 ms·

> Even being as charitable as I can, it just seems that you don't like the data because it doesn't fit your narrative. And you seem to like the data because it

> Even being as charitable as I can, it just seems that you don't like the data because it doesn't fit your narrative. And you seem to like the data because it does. You are right, I don't trust reported numbers of trans people hate crimes, because I'm not aware of any unified reporting standards around trans hate crimes (in general I don't trust reported numbers on newly reported hate crimes, especially when advocacy groups are doing most of the reporting; it means the issue is poorly understood (so badly reported) and highly politicized). I have a pretty large prior here that I believe trans hate crimes are underreported, much like I have a prior that sexual violence is underreported, due to the nature of these instances. Moreover, numbers this low have large uncertainty bands, just using basic frequentist or Bayesian probability methods, enough that I doubt we can even come to much of a conclusion over our topic of discussion. I think we'll have to agree to disagree, and please stop downvoting me. I felt like our discussion was productive.

9wzYQbTYsAIc 5y ago
> I felt like our discussion was productive. Possibly violating HN site guidelines here, but I wanted to add a civilly-toned comment to thank you for illuminating one area where our mental models had diverged on this topic.
dvt 5y ago
> And you seem to like the data because it does. I don't have a narrative. I look at the data and draw conclusions. You start with conclusions and try to morph the data to fit them. I'll even grant you that sexual crimes go underreported (heck, have 43% to bring up that number up), but even so, it wouldn't account for a "disproportionate" number of crime against trans people. It would barely equal the rate of the general population. You're seriously trying to argue that trans crimes are underreported by multiple factors? That's quite the claim. > I think we'll have to agree to disagree, and please stop downvoting me. FYI, you can't downvote direct children on HN.
- 9wzYQbTYsAIc 5y ago
  > It would barely equal the rate of the general population. You are generalizing from the specific, here. The data that you appear to be referencing is only looking at murder rates. You’ve then set up a straw-man argument. The straw is in the data that you’ve not incorporated into your mental model: the rates of crimes other than murder.
  - dvt 5y ago
    Yeah, I used murder rates specifically for three reasons: (1) they are often cited in news articles, including my citation above; (2) they are the easiest to compare side-by-site in an apples-to-apples comparison (gen pop vs population X or population Y); and (3) murder rates tend to be a good indicator of other, proximate, criminal activity (be it sexual assault, physical assault, etc.).
    - 9wzYQbTYsAIc 5y ago
      Citations for your assertion in 3? Also, your methodological techniques are faulty - ease of comparison is not a means in which to determine the validity of a comparison.
      - dvt 5y ago
        > Citations for your assertion in 3? https://www.pewresearch.org/fact-tank/2020/11/20/facts-about-crime-in-the-u-s/ft_20-11-12_crimeintheus_2/ https://www.pewresearch.org/fact-tank/2020/11/20/facts-about... > Also, your methodological techniques are faulty - ease of comparison is not a means in which to determine the validity of a comparison. This is a straw-man, nowhere do I argue that "ease of comparison is a means in which to determine the validity of a comparison." I do, however, argue that the comparison of murder rates in gen pop to the murder rates of trans pop is a valid comparison. Perhaps you'd like to elucidate why you don't think it is.
        9wzYQbTYsAIc 5y ago
        I’m not seeing where “U.S. violent and property crime rate have plunged since 1990s, regardless of data source” that you cited shows evidence that murder rates are reliable predictors of rates of other types of victim-having crimes? Additionally, pointing out errors of generalization and particularization is not a straw-man tactic of argument. It is a direct conflict with your argument on the grounds of scientific validity - are you actually measuring what you think you are measuring? Setting aside that we haven’t yet established that murder rates are reliable predictors of other victim-having crimes, there is a deeper problem, that another commenter was trying to point out to you in terms of Bayesian probability. To draw the conclusion that I drew, which I continue to defend (for clarities sake: that the LGBTQIA+ community is disproportionately affected by victim-having crime), you would need to look at data that partitions the general public into its various sub-groups and compares the frequency of victim-having crime between all of the individual sub-group, in all combinations. Such a statistical technique is frequently used in empirical studies of populations, across disciplines. Statistical analysis of these between-group variations is where you are able to draw out conclusions such as “blacks are disproportionally convicted of certain crimes” or “LGBTQIA+ are disproportionately victims of murder, physical abuse, emotional abuse, and widespread discrimination”. > Perhaps you’d like to elucidate why you don’t think it is [a valid comparison] There may very well be some valid conclusions to draw from data about murder rates for the whole population compared to a given sub-group, but that methodological technique suffers from the so-called “law of averages”. See the Expectation values example in https://en.m.wikipedia.org/wiki/Law_of_averages https://en.m.wikipedia.org/wiki/Law_of_averages, for example. In the scientific community, it is common knowledge that research methodologies based on statistical analyses of between-group variations is a technique that is sensitive to patterns that would not appear in a comparison with the average.
- Karrot_Kream 5y ago
  If you'd prefer working with the data we have presented, take a look at https://ucr.fbi.gov/hate-crime/2019/tables/table-1.xls https://ucr.fbi.gov/hate-crime/2019/tables/table-1.xls . Other than racially motivated hate crimes which comprise most FBI recognized hate crimes, the next most is religious hate crime, below which is sexual orientation motivated hate crimes. And sexual orientation motivated hate crimes rank very similarly to religiously motivated hate crimes. In other words, sexual orientation oriented hate crimes are the 3rd most frequent, and very close in # to religiously motivated hate crimes, the 2nd most frequent hate crime. Specifically with anti-gender-identity based hate crime, we can see that anti-transgender hate crime has one of the highest incident numbers for any individual cohort, despite knowing that only 0.6% of the population identifies as trans according to data presented earlier in this thread.
  - 9wzYQbTYsAIc 5y ago
    This is well-reasoned statistical thinking with an obvious basis in sound research methodology and does not deserve to be grayed out.

9wzYQbTYsAIc 5y ago

> I felt like our discussion was productive. Possibly violating HN site guidelines here, but I wanted to add a civilly-toned comment to thank you for illuminating one area where our mental models had diverged on this topic.

dvt 5y ago

> And you seem to like the data because it does. I don't have a narrative. I look at the data and draw conclusions. You start with conclusions and try to morph the data to fit them. I'll even grant you that sexual crimes go underreported (heck, have 43% to bring up that number up), but even so, it wouldn't account for a "disproportionate" number of crime against trans people. It would barely equal the rate of the general population. You're seriously trying to argue that trans crimes are underreported by multiple factors? That's quite the claim. > I think we'll have to agree to disagree, and please stop downvoting me. FYI, you can't downvote direct children on HN.

9wzYQbTYsAIc 5y ago
> It would barely equal the rate of the general population. You are generalizing from the specific, here. The data that you appear to be referencing is only looking at murder rates. You’ve then set up a straw-man argument. The straw is in the data that you’ve not incorporated into your mental model: the rates of crimes other than murder.
- dvt 5y ago
  Yeah, I used murder rates specifically for three reasons: (1) they are often cited in news articles, including my citation above; (2) they are the easiest to compare side-by-site in an apples-to-apples comparison (gen pop vs population X or population Y); and (3) murder rates tend to be a good indicator of other, proximate, criminal activity (be it sexual assault, physical assault, etc.).
  - 9wzYQbTYsAIc 5y ago
    Citations for your assertion in 3? Also, your methodological techniques are faulty - ease of comparison is not a means in which to determine the validity of a comparison.
    - dvt 5y ago
      > Citations for your assertion in 3? https://www.pewresearch.org/fact-tank/2020/11/20/facts-about-crime-in-the-u-s/ft_20-11-12_crimeintheus_2/ https://www.pewresearch.org/fact-tank/2020/11/20/facts-about... > Also, your methodological techniques are faulty - ease of comparison is not a means in which to determine the validity of a comparison. This is a straw-man, nowhere do I argue that "ease of comparison is a means in which to determine the validity of a comparison." I do, however, argue that the comparison of murder rates in gen pop to the murder rates of trans pop is a valid comparison. Perhaps you'd like to elucidate why you don't think it is.
      - 9wzYQbTYsAIc 5y ago
        I’m not seeing where “U.S. violent and property crime rate have plunged since 1990s, regardless of data source” that you cited shows evidence that murder rates are reliable predictors of rates of other types of victim-having crimes? Additionally, pointing out errors of generalization and particularization is not a straw-man tactic of argument. It is a direct conflict with your argument on the grounds of scientific validity - are you actually measuring what you think you are measuring? Setting aside that we haven’t yet established that murder rates are reliable predictors of other victim-having crimes, there is a deeper problem, that another commenter was trying to point out to you in terms of Bayesian probability. To draw the conclusion that I drew, which I continue to defend (for clarities sake: that the LGBTQIA+ community is disproportionately affected by victim-having crime), you would need to look at data that partitions the general public into its various sub-groups and compares the frequency of victim-having crime between all of the individual sub-group, in all combinations. Such a statistical technique is frequently used in empirical studies of populations, across disciplines. Statistical analysis of these between-group variations is where you are able to draw out conclusions such as “blacks are disproportionally convicted of certain crimes” or “LGBTQIA+ are disproportionately victims of murder, physical abuse, emotional abuse, and widespread discrimination”. > Perhaps you’d like to elucidate why you don’t think it is [a valid comparison] There may very well be some valid conclusions to draw from data about murder rates for the whole population compared to a given sub-group, but that methodological technique suffers from the so-called “law of averages”. See the Expectation values example in https://en.m.wikipedia.org/wiki/Law_of_averages https://en.m.wikipedia.org/wiki/Law_of_averages, for example. In the scientific community, it is common knowledge that research methodologies based on statistical analyses of between-group variations is a technique that is sensitive to patterns that would not appear in a comparison with the average.
Karrot_Kream 5y ago
If you'd prefer working with the data we have presented, take a look at https://ucr.fbi.gov/hate-crime/2019/tables/table-1.xls https://ucr.fbi.gov/hate-crime/2019/tables/table-1.xls . Other than racially motivated hate crimes which comprise most FBI recognized hate crimes, the next most is religious hate crime, below which is sexual orientation motivated hate crimes. And sexual orientation motivated hate crimes rank very similarly to religiously motivated hate crimes. In other words, sexual orientation oriented hate crimes are the 3rd most frequent, and very close in # to religiously motivated hate crimes, the 2nd most frequent hate crime. Specifically with anti-gender-identity based hate crime, we can see that anti-transgender hate crime has one of the highest incident numbers for any individual cohort, despite knowing that only 0.6% of the population identifies as trans according to data presented earlier in this thread.
- 9wzYQbTYsAIc 5y ago
  This is well-reasoned statistical thinking with an obvious basis in sound research methodology and does not deserve to be grayed out.