It sounds like you're suggesting that people are making comments along the lines of "This poll must be wrong because all the black people I know think police presence should be reduced."
I scanned the first several posts and didn't see any comments like that. Someone mentioning their personal experience doesn't align with a result isn't suggesting the poll is wrong. Not everything is an argument. To me this suggests the data points could be distributed into homogeneous pockets, and a discussion of this possibility seems fair game.
Some comments raise questions about the experimental methods or muse about hidden nuance in the data, which also seem like fair discussion points for this forum.
This is fascinating because I am myself half black, lived in many communities of mixed demographics and never once heard a friend (of any race) say they want more police at all, ever.
Depending on people’s age, it would be very surprising to hear anyone advocate for more police. Police generally are seen as the folk who take the fun out of things. So until you start getting affected by their absence you don’t advocate for them.
I’ve lived in many kinds of neighborhoods, from one where there were six murders in a school year to one where there wasn’t one in living memory. At one time I was poor enough that taking the bus was a luxury and I’d rather walk and save the change. I’ve met and known lots of kinds of people.
I agree that comment uses anecdotal data. However, I don't think that it is attempting to refute the findings of the original post.
Anecdotal data isn't inherently bad; it just shouldn't be used in attempt to prove or disprove anything, which I don't think that comment does. Using anecdotal data to suggest the possibility of alternative or more-nuanced hypotheses in an informal discussion thread is a totally appropriate use of it.
Furthermore, this is only one example. Drawing conclusions about a trend from a single example is ironically what the original commenter said is not "smart."
I scanned the first several posts and didn't see any comments like that. Someone mentioning their personal experience doesn't align with a result isn't suggesting the poll is wrong. Not everything is an argument. To me this suggests the data points could be distributed into homogeneous pockets, and a discussion of this possibility seems fair game.
Some comments raise questions about the experimental methods or muse about hidden nuance in the data, which also seem like fair discussion points for this forum.