Wikipedia:Reference desk/Archives/Science/2016 September 23

= September 23 =

Child sex abuse statistics
Recently, I volunteered for a program working with children. As part of the required training everyone had to review a set of materials discussing child sex abuse. In those materials were a disturbing set of unreferenced statistics. I'd like to figure out if the statistics presented were accurate / reliable or cherry-picked for dramatic effect. Claims:


 * 1) Only 3% of the perpetrators of child sexual abuse are ever identified and prosecuted.
 * 2) Over a lifetime, a perpetrator who isn't caught can victimize up to 500 children.

So far my google-fu hasn't been able verify or source either claim. I'm sure child sex abuse is underreported, but even so, 3% seems jaw-droppingly low. The 500 victims claim (7 / year for 70 years?) seems like an extreme worst case scenario, if it isn't simply an example of magical thinking. Can anyone help figure out where the claims might have come from and whether they are accurate? This is for a US context, if it helps. Dragons flight (talk) 09:46, 23 September 2016 (UTC)


 * This says "The belief that child sexual abusers are incorrigible recidivists is also an oversimplication. In reality, the overall re-offense rate for child molesters is lower than that for other criminals." Refs 47-49 there also seem like they may have statistics on victimization counts. This  looks to be a very review good article, summarizing many key research studies and statistics. SemanticMantis (talk) 13:05, 23 September 2016 (UTC)


 * 1) You will need to narrowly define "child sex abuse". Public urination, seen by a child, could be included, and men have ended up on the watch list because of it.  Consensual sex between an 18 year-old male and a girl just below the age of consent can have the same result.  Presumably, a large portion of such cases are never charged, especially if the genders are reversed, as the punishment seems wildly out-of-whack with the crime.


 * 2) Note that any "statistic" containing the words "up to" means there was a single case, no claim beyond that was made. StuRat (talk) 13:20, 23 September 2016 (UTC)


 * Regarding the empirical worst-case scenario of the "up to" wording, the lead of Jimmy Savile sexual abuse scandal says:


 * The Metropolitan Police stated that the total number of alleged victims was 589, of whom 450 alleged abuse by Savile.[4][5]


 * Loraof (talk) 13:30, 23 September 2016 (UTC)
 * With this scale of horror, it seems odd to nitpick, but of that 450, "only" 80% were "children or young people" at the time, which is around 360. Smurrayinchester 13:51, 23 September 2016 (UTC)


 * Well, the question has basically been answered, but in any case those two stats rise red flags for the concerned statistician. If "3% of the perpetrators (...) are ever identified and prosecuted", one absolutely needs to ask how that number was obtained, since by definition it is a figure that is hard to know (it is possible to estimate it with various methods, but never accept such claims without knowing what degree of credibility the methodology has).
 * The second one is an "up to" statistic; while it produces a significant emotional impact, it is of little relevance in that case (since that is a big statistical outlier). "Up to" stats can be useful, e.g. you want the bridge to hold "up to X kg", because even if traffic is usually low, you are not satisfied with one collapse per year; but more often than not, they are not the correct measure. That is of course why websites report "up to" X simultaneous visitors, etc. (And of course, a perpetrator "can" rape up to 100,000 victims if starting young enough as a despot, but "can" is the operative word.)
 * Of course, child abuse is still a big problem, but it does not justify abusing the statistics. I remember than for our driving test we were fed statistics about road deaths, and although some good ones would already have been impressive enough, they were often phrased in misleading way to make an even worse impression. E.g. "x% of road accidents happen on commuting", to tell you to pay attention even on the roads you know well, but of course the relevant number is the deaths by kilometer travelled (or per hour travelled, maybe), which could be either higher or smaller on commuting trips. Tigraan Click here to contact me 16:15, 23 September 2016 (UTC)


 * The worse stats was "alcohol related car deaths", which could mean anyone in any vehicle involved had been drinking, even a passenger in a car that was hit, versus "alcohol caused", which is the relevant stat. StuRat (talk) 16:18, 23 September 2016 (UTC)
 * 1. The claim rests on a source of "identifying" perpetrators. Such a source might be Child protection authorities who might see a ratio of 33 suspects who merit on-going investigation to every actual prosecution, or it might be a Mental health authority who observes a number of cases considered at risk. In either case the data source would be confidential and unapproachable.
 * 2. The claim contains the weasel wording "up to" 500 that makes the claim almost meaningless, except for its shock value, and almost incontrovertable. Its illusory numerical accuracy leads only to a useless speculation whether exhaustion alone would limit a lifelong (say age 18 to 80) dedicated child molester to victimize (in ways unspecified) a new child on average every 45 days. A source might defend the claim by invoking a concept of collective victimization in a special environment such as a school, Seminary or enslaved area where Child sexual abuse was prevalent. AllBestFaith (talk) 16:49, 23 September 2016 (UTC)

Can being fat ever be healthy?
People always say that being fat is unhealthy. But how fat does a person have to be to receive unflattering comments? Is it ever possible that a person just happens to be born with an undesirable body shape but manages a very balanced diet and maintains regular physical activity with no intention of changing the figure? Can such a fat but able person be possible and healthy? 64.134.39.172 (talk) 19:23, 23 September 2016 (UTC)


 * The answer may well depend on how fat the person is. If morbidly obese, then probably no, that can't be healthy.  However, if just slightly overweight, then it's possible, especially since some extra weight is protective against some diseases that cause sudden weight loss.  Not that it prevents the disease itself, mind you, but just that it prevents weight, blood sugar level, etc., from dropping to a dangerously low level.  On the other side of the equation, susceptibility to heart disease, diabetes, and other medical problems also has a genetic component, so if someone is lucky enough to be resistant to those conditions, those few extra pounds are unlikely to be a problem.


 * As far as unflattering comments go, fashion models have set unrealistic expectations, and many of them are underweight, and/or their photos are altered so they look that way. So, a person may very well be normal weight, and yet be criticized for being fat.   StuRat (talk) 19:29, 23 September 2016 (UTC)


 * Sumo wrestlers seem to have a very big figure. Are they considered more unhealthy than the rest of Japan's population? 64.134.39.172 (talk) 19:55, 23 September 2016 (UTC)


 * If this is a serious question you should be asking your doctor or nutritionist, not for our opinion on the subject. And where did the notion that the rest of Japan's population is unhealthy come from? μηδείς (talk) 20:31, 23 September 2016 (UTC)


 * From our article: "Sumo wrestlers have a life expectancy between 60 and 65, more than 10 years shorter than the average Japanese male. Many develop diabetes or high blood pressure, and they are prone to heart attacks due to the enormous amount of fat they accumulate." The question was not about the OP's own health, neither did it imply that the Japanese population are unhealthy. Tevildo (talk) 21:15, 23 September 2016 (UTC)


 * Well, like I said, if his question was serious. But try changing the word "fat" in his second sentence: "But how fat does a person have to be to receive unflattering comments?" to how effeminate, devout, short, or unable to tan, and get back to me on whether this is serious, or just trolling. μηδείς (talk) 01:56, 24 September 2016 (UTC)


 * Well, my doctor prefers that I should be slightly overweight rather than at my ideal BMI. So it is horses for courses. Greglocock (talk) 23:17, 23 September 2016 (UTC)


 * You can't be healthy and fat; if you stick to a healthy diet and get plenty of exercise and are nevertheless overweight, then you suffer from a medical condition like e.g. a thyroid problem. You have to consider this from the perspective of a hunter gatherer population, because our bodies evolved as hunter gatherers. Obviously becoming overweight on a normal calorie intake would be quite disastrous for a hunter gatherer, you would struggle to keep up with you fellow hunters when running down an antelope in the tropical heat. So, you would be of little use to society, you would be considered to be a handicapped person. Count Iblis (talk) 06:21, 24 September 2016 (UTC)


 * You are assuming everyone hunted. Don't forget those who gathered food and firewood, stayed home taking care of children, made weapons and tools, etc.  There was a division of labor quite early on. StuRat (talk) 15:21, 24 September 2016 (UTC)


 * Yes, but I think you'll struggle to find obese people in the hadza community :) . Count Iblis (talk) 18:21, 24 September 2016 (UTC)


 * It's now very clear that, at least in wealthy western nations, people in the BMI range considered "overweight" tend to live longer than those in the so-called "healthy" weight range. There's a lay summary of this issue here.  That source says that one study found that among Italian women, the longest life expectancy was for those with a BMI of 33, which is normally considered "obese". Adrian J. Hunter(talk•contribs) 08:30, 25 September 2016 (UTC)


 * The "unflattering comments" is a bullying thing, not a science thing. There was a fairly long period of time when smokers were known to be more unhealthy, without any possible genetic excuse, but received few if any negative commentary.  And of course other non-harmful characteristics like skin color get bullying comments in some places.  It is clear that extremes of weight have a negative impact on health, but so do many other things, and usually, people don't take them to be their business.  As such, I have seen some news articles to suggest that Japan is more prone to such bullying (and, I suspect, other kinds as well).  As for the science, it looks like Adrian beat me to the relevant reference, so I'll leave it there. Wnt (talk) 16:19, 26 September 2016 (UTC)

1036 Ganymed
What is the composition, temperature, and pressure at the center of 1036 Ganymed?144.35.45.81 (talk) 20:06, 23 September 2016 (UTC)
 * It is irregularly shaped. So, it is difficult to pinpoint its center, which makes the composition uncertain as the asteroid is likely inhomogeneous. As to the temperature it likely to be the same as the average surface temperature: ~ 160K. The pressure is probably low: ~$$G\rho^2R^2\sim0.15$$ bar assuming &rho;~1 g/cm3. Ruslik_ Zero 20:36, 23 September 2016 (UTC)
 * The center is where the string points when you try to hang it in a large gravitational field. Now the center of a wok... Sagittarian Milky Way (talk) 22:41, 23 September 2016 (UTC)


 * I asked a somewhat related question at https://en.wikipedia.org/wiki/Wikipedia:Reference_desk/Archives/Science/2015_March_6#Digging_into_Vesta Wnt (talk) 16:22, 26 September 2016 (UTC)

WR 102
Currently the article on WR 102 has inconsistency, saying it's in Cygnus and elsewhere that it's in Sagittarius (and is also included in the Sagittarius template). I couldn't google a reliable source on that, but this book says it's in Cygnus. Could someone verify and fix this? Thanks. Brandmeistertalk  22:48, 23 September 2016 (UTC)
 * An online astronomical object catalog I checked says it's at 17 hours 45 minutes 47.56 seconds and negative 26 degrees 10 minutes 26.9 seconds. Cygnus is very far from there. Sagittarian Milky Way (talk) 23:05, 23 September 2016 (UTC)
 * Those co-ordinates are very close to the border of Sagittarius and Ophiuchus: we need to double check the co-ordinates given in the sources, and those of the border at present (precession and all that) carefully before making the necessary amendment.
 * [Update – as of Epoch 2000 (i.e. from Tirion & Sinnot's Sky Atlas 2000.0 2nd Ed 1998) they are just in Sagittarius, so I have corrected the text. I can't comment on Brandmeister's linked book claiming Cygnus, as I can't read Russian.]
 * However, I notice that these co-ordinates are in the Article cited to its reference 1, a paper which gives them for its Star 2 in Table 5, but I see nothing in that paper to confirm that this star is WR102. I don't suppose that it isn't, but we do perhaps need an explicit confirmation of the identity of the two. {The poster formerly known as 87.81.230.195} 90.202.211.191 (talk) 23:34, 23 September 2016 (UTC)