But how is my anecdote any less relevant than the others above, which were in fact deemed acceptable commentary on the post? Indeed, you seem to be arguing against the thrust of the post itself, rather than my individual comment.
Also, does mere failure of an example warrant downvoting? (Rhetorical question; it doesn’t. For instance, it may prompt a discussion of the reasons the example fails, which may add value to the discussion. Or it may be neutral. Downvoting implies the added value is outright negative.)
I’m not happy with many of the other anecdotes either. Example being misleading is a negative effect. Seeing patterns where there’s none is a basic failure in human mind, we should know better than to indulge it.
(1) The existence of the phenomenon under discussion was not itself in dispute;
(2) The failed example is likely to lead to confusion about the nature of the (undisputed) phenomenon;
and
(3) The commenter should have known this, or in some other way the example was not offered in good faith.
Looking through your previous comments, and noting your citation of Gelman, it appears that you are skeptical of the name-bias phenomenon itself as posited by Yvain and those he cites; hence (1) does not hold. Regarding (2), I find it highly unlikely that someone following the discussion about whether name-bias exists is going to be misled or confused by a particular anecdote that may or may not exemplify the phenomenon (depending in particular on whether the phenomenon exists). As for (3), the comment was quite consistent with the spirit of the preceding discussion. It should also be noted that (as I hinted) the name coincidence in the court decision would not have caught my attention if I hadn’t read the post—this fact alone arguably makes it worthy of a comment on said post.
It would be different if you accepted Yvain’s claims (1), but thought my example would mislead folks about what it was that Yvain was describing (2). As it is, however, a court decision such as I mentioned clearly counts as a candidate for name-bias, if you assume that name-bias as described in the post actually exists. The issue is that you don’t accept this premise in the first place. Consequently I don’t see a defensible rationale for downvoting in this context. (Certainly not without also downvoting the main post.)
I accept that the effect might be real, but neither the main post, nor the paper itself are at odds with interpreting this effect as tiny. Being tiny, the effect can’t be used in interpreting the structure of specific examples, it doesn’t suggest that any given example has good chances of being a name-based decision.
But how is my anecdote any less relevant than the others above, which were in fact deemed acceptable commentary on the post? Indeed, you seem to be arguing against the thrust of the post itself, rather than my individual comment.
Also, does mere failure of an example warrant downvoting? (Rhetorical question; it doesn’t. For instance, it may prompt a discussion of the reasons the example fails, which may add value to the discussion. Or it may be neutral. Downvoting implies the added value is outright negative.)
I’m not happy with many of the other anecdotes either. Example being misleading is a negative effect. Seeing patterns where there’s none is a basic failure in human mind, we should know better than to indulge it.
I could see a case for this if:
(1) The existence of the phenomenon under discussion was not itself in dispute;
(2) The failed example is likely to lead to confusion about the nature of the (undisputed) phenomenon;
and
(3) The commenter should have known this, or in some other way the example was not offered in good faith.
Looking through your previous comments, and noting your citation of Gelman, it appears that you are skeptical of the name-bias phenomenon itself as posited by Yvain and those he cites; hence (1) does not hold. Regarding (2), I find it highly unlikely that someone following the discussion about whether name-bias exists is going to be misled or confused by a particular anecdote that may or may not exemplify the phenomenon (depending in particular on whether the phenomenon exists). As for (3), the comment was quite consistent with the spirit of the preceding discussion. It should also be noted that (as I hinted) the name coincidence in the court decision would not have caught my attention if I hadn’t read the post—this fact alone arguably makes it worthy of a comment on said post.
It would be different if you accepted Yvain’s claims (1), but thought my example would mislead folks about what it was that Yvain was describing (2). As it is, however, a court decision such as I mentioned clearly counts as a candidate for name-bias, if you assume that name-bias as described in the post actually exists. The issue is that you don’t accept this premise in the first place. Consequently I don’t see a defensible rationale for downvoting in this context. (Certainly not without also downvoting the main post.)
I accept that the effect might be real, but neither the main post, nor the paper itself are at odds with interpreting this effect as tiny. Being tiny, the effect can’t be used in interpreting the structure of specific examples, it doesn’t suggest that any given example has good chances of being a name-based decision.