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”.
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.
> 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?
I'm not sure if you're being purposefully obtuse here, but there's an obvious correlation between the drop in "violent crime" (homicide, murder, assault, manslaughter, etc.) and "violent victimization" (physical abuse, sexual abuse, verbal abuse, etc.). Do I have to calculate the correlation coefficient for you?
> 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.
I don't believe the comparison here suffers from the law of averages. And even if it did, the difference between the two data sets' deviation isn't high enough to be of any significance.
Well just have to agree to disagree, but I’ll refer you to the empirical evidence cited by myself and other commenters that ultimately supports my original assertion.
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, 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.