This article is within the scope of WikiProject Insects, a collaborative effort to improve the coverage of insects on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
The amount of vandalism that keeps appearing on this article is ridiculous. I suggest that this article should be semi-protected to prevent IP users from vandalising. Burklemore1 (talk) 09:18, 13 October 2015 (UTC)
Sounds reasonable, but I don't know how to nominate an article for protection. Pchown (talk) 22:20, 14 October 2015 (UTC)
I updated the discussion of mosquito size using figures from Service, Medical Entomology for Students.
The old figure for mosquito weights was inaccurate I think, but I don't have a very good source. In Keith Snow, Mosquitoes, there is a graph showing weights of individuals from a population of Aedes punctor. The midpoint of the graph is about 5mg so I used that in the article, because it's likely to be more accurate than the old figure (up to 2.5mg, which is clearly contradicted by the graph). Does anyone have a better figure, though?
What eats mosquitoes besides the Toxorhynchites larva, the only mentioned predator? Birds? Other insects, like praying mantis, maybe? (I searched the article for "eat", "feed", "prey", "consume" without finding out.) Thank you, Wordreader (talk) 05:11, 29 January 2016 (UTC)
The PlosONE article cited in support of the idea that attractiveness to mosquitoes has a heritable component is very weak evidence. It consists of poor statistical analysis supporting a correlation that appears to be rather weak even if true. Verytas (talk) 13:24, 22 May 2016 (UTC)
Well, tweak the entry to say "provides weak support for" or something of that sort. Chiswick Chap (talk) 15:28, 22 May 2016 (UTC)
Verytas: I'd be interested to know what the concern is. Is it the small sample size? I looked at the comments posted on the Plos One site, and none of them are critical of the study, which seems surprising if there is a significant weakness. Pchown (talk) 15:35, 22 May 2016 (UTC)
I call this "shotgun science". The datapoints in the graphs look like the result of shotgun hit, not a true correlation. Indeed, you can almost always get an apparent correlation on this type of graph from a random distribution. Sample size is not the issue. Look at the central figure in this paper illustrating correlations - most values on one axis include most values on the other axis. See also Anscombe's Quartet for an illustration of the weakness of this kind of data. Verytas (talk) 18:26, 22 May 2016 (UTC)
Why not start by posting a comment on the Plos One site, and see what the authors have to say? It feels a bit like WP:Original_Research if we're negative about a paper here, when there don't seem to be negative comments in the scientific literature. Unfortunately I also don't understand your reasoning about it not being a 'true' correlation. There is a mathematical test for correlation which is met by the data the authors present. Looking at the graph, to me it looks correlated; that's subjective of course but Anscombe seemed to be arguing for the need to take a subjective look at data. Pchown (talk) 19:11, 22 May 2016 (UTC)
Actually I've just looked again and I see what you mean about the relative attraction graph. Visually the correlation doesn't look that strong. Why not ask the authors about it? I'd be interested to know what they say. Pchown (talk) 19:24, 22 May 2016 (UTC)