Changes

:Medline, one of the largest databases relating to medical articles, returned just two articles for a search of ""Breast Neoplasms (exp)" and "Hollywood". One of these only referenced Hollywood because to establish a timeframe (it was translated from Japanese, and used the term "Hollywood syndrome" in its title) whereas the other referred to the geographic location of a clinic. I think this establishes that there is no medical evidence supporting the claim that Hollywood somehow increases the likelihood of breast cancer. If you have alternative search terms or databases you want me to use post them here and I will. [[User:StatsMsn|StatsMsn]] 22:46, 3 May 2008 (EDT)

:::Medline reports from almost all major medical journals, it does not discriminate based on political ideology. All you have done is listed a few cases and suggested that being famous and connected to Hollywood somehow increases the risk of breast cancer. You have not provided any evidence that the average age of breast cancer amongst stars is any less than the broader average. It is an unfortunate fact that young people do get breast cancer, but there is nothing to suggest that rate of young people in Hollywood being diagnosed with breast cancer is any different to the rate of young people in broader society. Until you refute this point you cannot claim that there is a mystery. [[User:StatsMsn|StatsMsn]] 23:26, 3 May 2008 (EDT)

When are you going to realise that you cannot base statistical conclusions upon a non-random sample? You are selecting a category of people who are by definition, far, far younger than the normal population (actresses and pop stars being typically in an age range from 18-45). Therefore, any incidences of breast cancer you find within that population will of course appear to trend younger. Please, take Statistics 101 before jumping to such ludicrous conclusions - you're simply making a fool of yourself. [[User:Bongabill|Bongabill]] 14:19, 4 May 2008 (EDT)

Sure it can be compared Andy, but you haven't done that. If you compare the number of young Hollywood breast cancer victims (of course you might want to define what you mean by that first) to the rest of the population in the same age group, fine. You haven't done that. You might also want to clarify where you got your 500 from - clearly an estimate, but based on what? [[User:Murray|Murray]] 17:47, 4 May 2008 (EDT)

::Mr.Schalfly, actually, I am a professional statistician with both graduate and post-graduate degrees in the subject, and also a 34 year professional career as a statistician working for the official statistics bureau of a major nation. Now, ''of course'' you can compare the 'Hollywood' 18-45 population to the general 18-45 population, but first you have to define that population. How are you defining ''your'' population? Is it that they are living in "Hollywood" - in which case it might be interesting to examine geographical factors upon the incidence of disease? No - as I see you are including Australians who live in the UK, people born in six different countries, and it seems none of the members of your data set are actually ''from'' the geographical region of ''Hollywood'', and in fact many of them do not even live there either. So how are you defining members of your data set?

::Secondly, I note that your article does not compare your data set to the general 18-45 population, but instead compares it to the general population of 'all ages', quoting an ''average'' age of diagnosis of 64 - which is clearly outside your terms of reference. I'm afraid there's absolutely nothing correct here with any of your reasoning, which not only fails Statistics 101, but fails it in the first semester. This article is specious nonsense. [[User:Bongabill|Bongabill]] 14:59, 4 May 2008 (EDT)

:::Just to point out, Mr. Schlafly, you fell back on your "record" of statistical courses as evidence to your statistical competency. [[User:AndrasK|AndrasK] 15:27, 4 May 2008 (EDT)

::::Actually ASchalfly, it was indeed you who first raised the issue of credentials in this thread - your comment "I doubt you've taken half the statistics courses I have" came after Bongabill's first and only comment on Conservapedia, and which never mentioned his or her credentials. [[User:StatsFan|StatsFan]] 18:52, 4 May 2008 (EDT)

All right Andy, I'll make a couple of suggestions. I suggest clarifying the basis of the estimate that there are 500 Hollywood women under 35 who would admit to having breast cancer. I don't understand where the number comes from - what was the method or source used to come up with 500? If that's a good estimate, your comparison is potentially legitimate. Then, I would suggest you do an actual statistical test to determine whether the numbers are significantly different. [[User:Murray|Murray]] 18:48, 4 May 2008 (EDT)

::I'm not disputing them, I'm asking where they came from. So you looked up how many albums and movies are typically released in a year and extrapolated from that? [[User:Murray|Murray]] 19:05, 4 May 2008 (EDT)

::: That provides a ballpark estimate. Given that the incidence in the [[Hollywood]] community is an order of magnitude larger than the general population, ballpark estimates are enough to reveal a problem.

::::Andy, my man, a "ballpark estimate" is one thing. Pulling numbers out y'arse is quite another. If this really means something to you, then go and do some ''proper'' research. Lord knows you've got enough time on your hands. Reliable numbers may or may not be forthcoming, but at least you'll have tried and you'll be able to argue from a much stronger position if you succeed. Meanwhile, I'd quietly retreat from this one if I were you.

I wonder how many of those who were included in the list of breast cancer cases would be included in the list of 500. If that 500 is based on albums and movies released, it appears that many of those women shouldn't be included. Shirley Temple is the only movie star there. Most appear to be television actresses, and I suspect many more fall into that category than movie actresses. So I suspect that 500 is a pretty big underestimate. Either way though, if you want to make the case, you should include a real statistical analysis. For example, in the last sentence, you could say ''There is likely no more than 1000 in the population of Hollywood types under age 35 who would have publicly disclose breast cancer. The list above has 8 cases (and growing). The expected rate is only 1.5 out of 1000 in that age group. A chi-square test of proportions indicates that the difference between these groups is not statistically significant, X² (1 df) = 3.62, p = .109.'' [[User:Murray|Murray]] 21:28, 4 May 2008 (EDT)

::Mainly because I think the numbers given in the article have no basis in reality. The 500 is said to be based on the number of movies and albums released each year. According to [http://dissertations.ub.rug.nl/FILES/faculties/feb/2007/s.a.delre/c5.pdf] there are 473 movies each year - I think it's safe to say that nearly all have multiple women who would be included in this list if they had the misfortune to get breast cancer. And taking a look at the list, it's clear that the article is also referring to television actresses, in the US and across the world. So I think 500 is a gross underestimate. Even if the true population of female "Hollywood types" under 30 was 1000 (and I suspect it's more than that), the difference is not significant. [[User:Murray|Murray]] 23:28, 4 May 2008 (EDT)

::::::Well, see, there's a couple of problems with that analysis. Let's assume for the moment that your guesstimate of 500 actresses/singers at any one time whose schedules would necessitate an illness disclosure is roughly accurate. Problems: 1)Your examples are not drawn from a single point in time, but from over a period of decades (which would stretch the 500 to thousands). 2)Many of these women, while well-known, were ill at a time in their life when disclosure would not have been necessary (that is, when they were no longer active, and not part of the 500). 3) Many of your examples are deeply obscure, and certainly not part of the 500.--[[User:RossC|RossC]] 08:59, 5 May 2008 (EDT)

Except there's just one thing - we know that's not true. So what's wrong with the reasoning? Selective sampling, that's what's wrong. And which is what you're doing in this article. [[User:Bongabill|Bongabill]] 14:33, 4 May 2008 (EDT)

:Can you point out logical discrepancies between your logic and his? Refusing to respond does not give you credit, it makes you seem even more wrong. If your argument is correct, then why can you not point out problems in his? And by the way, taunting someone regarding one simple spelling error is puerile. [[User:AndrasK|AndrasK]] 14:42, 4 May 2008 (EDT)

:::Andy to try and translate Bongabill's reasoning into simpler terms: '''Selection bias/selective sampling is where the sample you choose for your study is rigged to give an inaccurate result'''. For example, let us suppose that Planned Parenthood did a study into teens and promiscuity and ecluded teens with STDs and tried to use that study to show teens were having "safe sex". Obviously this would be false since the sample was not ''representative'' of the teenage population. '''Likewise your sample (the list) is inaccurate since it only includes "hollywood" women who ever got breast cancer'''. Also to ignore Bongabill's point and to point out his spelling mistakes smacks of a childish ''ad homniem'' attack, to repeat AndrasK.[[User:Luder|Luder]] 16:05, 4 May 2008 (EDT)

This article is utter nonsense. The reasoning is absurd. It is just laughable. So you can name twenty well-known women who have had breast cancer. I can name twenty women I know who have had breast cancer. Does this mean women of my acquaintance are particularly susceptible? No, it's just a meaningless series of cases, anecdotal evidence of nothing whatsoever. This article is the same. Conservapedia really is a parody of itself! [[User:Humblpi|Humblpi]] 17:20, 4 May 2008 (EDT)

::''The incidence of breast cancer in the [[Hollywood]] culture can and should be compared to its incidence in the overall population.'' So where is this done? It is simple (the number of Hollywood types that have breast cancer)/(the number of hollywood types) vs. (the number of the overall population that have breast cancer)/(the number of the overall population). You have failed to provide even the simple mean, neverless the variance in order to conduct a hypothesis test. [[User:DanielB|DanielB]] 18:38, 4 May 2008 (EDT)

:::DanielB, you're correct, but that's not all - you would also need to conduct a Chi-Squared Test on the result to see whether or not the variance in the data actually had any statistical significance. Different values are not necessarily statistically significant. There is no valid or relevant statistical analysis in this article, none whatsoever. [[User:StatsFan|StatsFan]] 18:44, 4 May 2008 (EDT)

::::: Vastlty different rates could be the result of the variance of your sample. Your sample size might be too small to draw conclusions or as StatsFan said it might not be significant. Andy you should research statistical hypothesis testing of Student and Fisher instead of being [[deliberate ignorance|deliberalty ignorant]] of them. [[User:DanielB|DanielB]] 19:03, 4 May 2008 (EDT)

::::::: What does by a factor of 2 mean? Do you mean a probabilty of less than 2% that the difference in means is a result of random sampling? As you have failed to establish a population size, how do you know the degrees of freedom of your t-distribution? You appear to have a vague understanding of statistic hypothesis testing without any actual working knowledge of how it works. [[User:DanielB|DanielB]] 20:42, 4 May 2008 (EDT)

:Based on [[Breast_cancer#Risk_factors|this]], perhaps [[Hollywood values]] encompass poor eating habits and abortions as well. [[User:Feebasfactor|Feebasfactor]] 13:18, 5 May 2008 (EDT)

:: From a factor of 2 to 5 to 10 times increase. These numbers are incredible and would be very impressive if you had any data what so ever to back this up. You have done no statistical analysis all you have done is make up "ballpark" numbers which suit your hypothesis. Get some data, analyse it and submit it to a medical journal if this as rock solid as you say. Otherwise there is no "mystery cause of Hollywood breast cancer" as you have failed to establish it is any higher than you might expect in the general population. [[User:DanielB|DanielB]] 02:01, 6 May 2008 (EDT)

:::: I am indeed serious (but don't fret--I get that a lot). I'll whip up a paragraph on possible causation when I have a bit more time (tomorrow, hopefully).--[[User:RossC|RossC]] 13:13, 6 May 2008 (EDT)

Your argument is hardly convincing.

::This does seem like a story that's not being covered in the MSM, and I think you've uncovered a real 'sleeper' news story with this original research. Do you have a press agent who can help you publicize it - can you get this into the news 'wires'? I'm sure various Hollywood press organs would be VERY interested in this story - why don't you try and get this to the LA Times, or Variety, or the Hollywood Reporter? This is Front Page material! [[User:RobCross|RobCross]] 14:07, 5 May 2008 (EDT)

Andy, maybe you should submit it to ''JPANDS''. [[User:Murray|Murray]] 11:00, 9 May 2008 (EDT)

[[User:AdenJ|AdenJ]] 23:41, 5 May 2008 (EDT)

Ahhhhh, again Aschlafly you dont know anything at all about me so I wouldn't throw around the liberal rubbish. Not that I have to prove anything to you however I will advise that in my countrys elections I vote for the right. Anyways my point is that this 'article' is not all convincing to anyone.

If there is any statistical significance of this, the math should be shown by the person making the claim. --[[User:Rutm|Rutm]] 14:34, 5 May 2008 (EDT)

:: Could you point to any study or news article akin to [http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2003/05/30/BA70131.DTL]? Have you ruled out other known risk factors? Where is the data to support this? I would be interested to see the data (rather than an list of 20 singers, actresses, and notables that at some time resided in Los Angeles) that supports the assertion. --[[User:Rutm|Rutm]] 16:11, 5 May 2008 (EDT)

== [[Liberal Denial]] ==

:Andy: (a) Hollywood isn't in Orange County. LA County is below the state average, according to Rutm's link. Even if it were higher, that says nothing about the rate in Hollywood, or among "Hollywood types". (b) Some of the comments above may be a bit strident. However, they raise challenges that haven't been answered. If you want to make this case, it would be better to show your methods and to actually do the math. The current approach is not convincing to those who understand statistics. [[User:Murray|Murray]] 21:47, 6 May 2008 (EDT)

::Andy the protest above are not about your conclusions they are about your compelete lack of data. What you are lacking is data. Rutm's observations are valid because they are based on data, something you lack. [[User:DanielB|DanielB]] 22:08, 6 May 2008 (EDT)

:::: Is it a small community? Do you have the exact number of people in this community or do you just have this strange 100-200 numbers from nowhere? You are making numbers up on gut instinct and calling it a sample. You would not get an arguement if you even had properly collected statistics.

The put-downs are on both sides, Andy ("emotional denial", "your minds are not open"). How about a compromise - if you will explain how you derived your sample size, a productive discussion can take place about whether or not this indicates an increased risk of breast cancer. Also, did you note above that the breast cancer rate in the county that Hollywood is in is actually lower than the state average? [[User:Murray|Murray]] 22:42, 6 May 2008 (EDT)

::I disagree that your comments are not put-downs in the same sense, but we can disagree there. I apologize for the comment. I pointed out that Hollywood is not in Orange County because your comment above used the rate in Orange County to support your argument. Sure, I will happily admit that it is ''possible'' that the breast cancer rate in the Hollywood district of Los Angeles is higher than in the general population. It would be a powerful statement if it could be shown, given that the rate in all of LA County is lower than the rest of the state. However, I am not convinced by the arguments currently put forth here. As I mentioned above, it would be helpful to explain your methods, particularly the method for determining the size of the sample. It would also be useful to define "Hollywood types", because it appears that many of those in the list, including 1 in the under 30 section, did not live in Hollywood. Research in any field is heavily dependent on explanations of how variables were defined and measured, and on descriptions of how the sample was derived and who it was composed of. [[User:Murray|Murray]] 23:15, 6 May 2008 (EDT)

== Reply to the above ==

:Andy, I interpreted your reference to the "small community" as referring to the town itself. Apparently I misinterpreted. Nobody is attempting to censor any discussion. I am in fact trying to have a discussion. I will ask again, would you please explain the methods that you used to derive these numbers, particularly the sample sizes but also the number of cancer diagnoses? If you are correct, there is nothing to lose, and you can only gain from being transparent. I have my doubts as to whether the population is truly as small as you say. However, if you can show that it is, then your hypothesis is correct. This is a matter of statistics, not of ideology. An intro stats text will go into detail about why it is important to be clear and detailed about methodology, and that is what is missing here.

If Z > Y, and passes the Chi Square Test, you're definitely onto something, and quite likely a large amount of funding would immediately be available for further research. But with A, B, W and X as yet undefined, as here, we don't have proof yet of the trend. But at least we have defined the numbers we are looking for, and we can continue to study your theory. [[User:StatsFan|StatsFan]] 17:57, 7 May 2008 (EDT)

::If you disagree with the methodology described above, what mathematical methodology do you instead propose? And what is this "fundamental reason" you refer to? I'm just curious as to how to prove this hypothesis. [[User:StatsFan|StatsFan]] 19:54, 7 May 2008 (EDT)

:::: The methodology allows other people to reproduce it and not require having an open mind or a statement from an authority. In making the claim "Hollywood types have higher breast cancer incidence than the general public" you are making a statement that can be tested. Thus, people want to know how you came to that conclusion and how to test it for themselves so they can see if it is true or not. --[[User:Rutm|Rutm]] 20:03, 7 May 2008 (EDT)

Andy, I have 2 questions for you. First, will you share your methods for determining the size of the sample and the number of cases? Second, what is your hypothesis as to causality? Thanks. [[User:Murray|Murray]] 20:35, 7 May 2008 (EDT)

::I take it you are refusing to share your methods, though it would be nice if you would just say that. ''We have the data that we have.'' True enough, but you're the only one who knows what it looks like. [[User:Murray|Murray]] 21:16, 7 May 2008 (EDT)

One other note - the point you make about "watch your smoker relatives drop dead" is an excellent example of selective sampling. I had two uncle's who smoked until they died in their nineties, neither of smoking-related illnesses. It doesn't prove for a moment though that smoking does not kill - it simply shows the effects of selective bias. [[User:StatsFan|StatsFan]] 20:59, 7 May 2008 (EDT)

::I am indeed confused by your metaphor. For me to know the answer to that I would need to know if it was common for a pitcher to throw three no-hitters, and if many Hall f Famers had thrown three no-hitters? If I'm correct (and I apologize if I am confused by your analogy), you are suggesting that since the data you present is 'obvious', then young Hollywood women do indeed contract breast cancer at higher rates, yes? [[User:StatsFan|StatsFan]] 21:24, 7 May 2008 (EDT)

::: StatsFan, the only info you need is the obvious one that no more than a few no-hitters occur each baseball season.

Then let's look at the data. Refusal to allow your conclusions to be examined (which is the effect of hiding your methods) suggests something to hide. In addition to knowing the frequency of no hitters, it would be important to know the number of games pitched. That's the piece that we're missing here. [[User:Murray|Murray]] 23:04, 7 May 2008 (EDT)

== Reply ==

:It's not legitimate to take those 100 without knowing how the number was determined. Why the reluctance to clarify? And the cause is completely orthogonal to the question of whether the rate is elevated. [[User:Murray|Murray]] 23:48, 8 May 2008 (EDT)

:It seems to me, (and forgive me if I'm wrong), but ''besides'' all of these young(ish) women being in show business (Hollywood-types) '''and''' being victims of breast cancer, thay all (must have) put their careers before the thought and "action" of having a family. It is a well known factor in risk assessment for BC that delaying childbirth or having an abortion has the likely but (of course) unsought side-effect of breast cancer. The woman so situated usually has the choice to advance her career '''or''' start her family but (often) the one comes at the expense of the other. [[User:MargeryCampbell|MargeryCampbell]] 12:45, 9 May 2008 (EDT)

Aschlafly, if what Margery said is what you actually meant then A) Why didnt you just state this a long time ago and avoid wasting your time and B) you just stated you learnt something from her which looks to me like you had no idea in the first place. Seems someone is straining to do some explaining.

== Reversion explained ==

:OK, but then what is the theory here? Because since most of these women did indeed have children, I can't see what the connections are to the theory? [[User:GaryK|GaryK]] 18:34, 9 May 2008 (EDT)

:::If it's that airtight, I would think it would be time to publish it somewhere other than here. Where are you going to send it? [[User:Murray|Murray]] 20:45, 9 May 2008 (EDT)

You know it occured to me Aschlalfy that when you are using the '100 or so singers' as an example you need to remove people like shirley temple from the list. You cant use her in relation to todays singers!

:::::Re. Margery Campbell's theory, I'm afraid that for your theory to be valid, the women listed here would have to be childless, which is not true in most of the cases, as I've pointed out above. In fact, the cases listed in this article could not seem to have ''less'' relevance to such a theory, and without a shadow of a doubt, the case of Soraya needs to be removed, as in her case, it was clearly hereditary. So what is the proposed theory that is so obvious? [[User:GaryK|GaryK]] 17:50, 9 May 2008 (EDT)

I'll repeat myself here also......Because until M.Campbell was involved you never mentioned this.

Since Schlafly has cut and run rather than try to back up his little pet thesis here, can we get this deleted now?--[[User:ChandaM|ChandaM]] 18:03, 17 May 2008 (EDT)

::So that's it, then, is it? You are going to ignore all the helpful statistical advice on the talk page and just let the groundless assertion in the article stand? Up to you. It's your website. [[User:Humblpi|Humblpi]] 05:45, 19 May 2008 (EDT)

[[User:AdenJ|AdenJ]] 05:47, 19 May 2008 (EDT)

::OK - here is my simple question. Where is the evidence to support the assertion that breast cancer incidence is higher among "Hollywood" women than in a general population of age-matched controls? [[User:Humblpi|Humblpi]] 13:27, 19 May 2008 (EDT)

::::Sorry. I asked for evidence, and you reply with nothing more than an assertion. I am quite prepared to believe that it might be so, but there is nothing to demonstrate whether it is or not. You present no evidence that the incidence is unusually high. How is the study population defined, and how big is it, and what is is the incidence of breast cancer? What is the control group, and how does the incidence there compare with that in the study population? And how do you control for biases such as reporting bias? It's an interesting idea, and deserves proper study, but without such a study you simply have no grounds for the assertion. [[User:Humblpi|Humblpi]] 14:02, 19 May 2008 (EDT)

:::::: Yes, an absolutely open mind. And yes, I have looked at the list. It is just that, a list. I could make a similar list of women of my acquaintance who have suffered from breast cancer, and it would produce an apparently high incidence (I won't right now go into the reasons why it would appear high, but I could do so if you really want to know) — but it would be meaningless. There is no defined study population, no defined control group, no stipulated method for gathering the data, no apparent attempt to take account of the biases inherent in compiling such a list. There is much detailed commentary on some of the statistical pitfalls and problems above, from various people, and I will not repeat it all here. What it comes down to is this: A statement that "Group X is particularly prone to develop problem Y" can only be made if the data are gathered by proper methods, and subjected to proper statistical analysis. You show neither adequate data collection nor adequate data analysis, and therefore have no evidence worthy of the name. I am quite open to the possibility that Hollywood women have a higher than normal incidence of cancer, and I can even think of some reasons why it might be so, but I see nothing to persuade me that it is so. The apparent high incidence looks much more like an artifact. [[User:Humblpi|Humblpi]] 14:50, 19 May 2008 (EDT)

:Andy, it wouldn't be possible to do a double-blind study. You can't randomly assign anyone to either be a rock star or to get a cancer diagnosis. I can't speak for anyone else I suppose, but I don't think anyone is asking for a formal study. The problem that keeps being raised is the estimate of the denominator, which appears (to me and apparently to others) to be a substantial underestimate; specifically, you haven't explained how you got the number. This is the point of Method sections in research papers - to explain how you did what it is that you did. Without that explanation, no matter how solid your case actually is, you leave yourself open to doubts. [[User:Murray|Murray]] 22:57, 19 May 2008 (EDT)

Well, you didn't ask me the question, so I didn't realize you wanted me to answer it, but ok. Before I do though, I will say that without a description of the method used, the reader is unable to make any informed judgment about whether there's cause for concern. That done, there are two things that I would want to see. First is whether the effect of Hollywood on breast cancer is statistically significant. Second, I would want to see an effect size. The first indicates the probability that any difference among Hollywood types is due to chance, and the second indicates how big the difference is. Both are needed. If there is a significant difference, but the effect size shows that the risk is (as an example) 2% higher, I wouldn't say that's particularly meaningful. Similarly, if the effect size shows that the risk is 100% higher but the p-value is .5 (ie, 50% chance that it's due to chance), I wouldn't make much of it. On the other hand, if it was significant and the effect size was on the order of a 10% increase in risk or more, I would think of that as meaningful. Notably, if your hypothesis is correct and there is a genuine effect, the effect size is likely to be closer to 10% higher than 100 or even 10x higher. Cancers are mostly determined by multiple risk factors so it's unlikely that one is going to account for all that much variance. [[User:Murray|Murray]] 12:32, 20 May 2008 (EDT)

::::I think the main problem is the assumption that stars who are diagnoed with breast cancer would try to cover it up. Unlike early stage Parkinson's disease, early stage breast cancer has a much smaller impact (if any) on the ability of a person to appear in a movie. Unlike Parkinson's diease (which cannot be cured) breast cancer can be treated, if the treatment is successful then the star can return to acting/singing, if not then signing a contract will be the least of their worries. So, using the example of Michael J. Fox really doesn't do anything to show that stars with breast cancer would cover it up, unless you've got some cold hard evidence of this then all of the statistics on this page are meaningless. [[User:StatsMsn|StatsMsn]] 06:26, 20 May 2008 (EDT)

Was that ("I don't see a substantive answer") addressed to me? I gave a quite clear answer. [[User:Murray|Murray]] 00:14, 21 May 2008 (EDT)

::ASchlafly, in the case of your hypothetical mine, the answer is an ''unreserved'' "of course not". But that is not a valid example, as in your story, you '''know''' the "number of women in the mine". In the case of young women in the entertainment business, an estimate of 50 is likely orders of magnitude short of reality - you don't "know the number of women in the mine". However, if you can figure out how many young women are in the entertainment business worldwide, you can contnue to test your thesis. But until then, simply stating "the number of women in the entertainment business worldwide is 50" is ''not'' an adequate answer. [[User:EngelUmpocker|EngelUmpocker]] 19:04, 22 May 2008 (EDT)

Andy, if you genuinely think that I didn't answer before you either didn't understand or didn't read what I wrote. I explicitly said a 10% increase and a sufficiently low p value. That couldn't be clearer. But fine, I'll do the research for you when I have the time, to clarify that 50 is a gross underestimate. [[User:Murray|Murray]] 19:44, 22 May 2008 (EDT)

::::The rational answer is that it depends on the disease, and the likelihood that it could be related to mining. Based on working in close quarters, 80% of the women working in the same mine could catch the same case of chicken pox, but no one would suggest that mining causes chicken pox. --[[User:DinsdaleP|DinsdaleP]] 21:26, 22 May 2008 (EDT)

:::::::No mater whether one has a 'problem' with Wikipedia or not, there's no denying these numbers - I personally know of at least half of them myself. [[User:EngelUmpocker|EngelUmpocker]] 20:08, 22 May 2008 (EDT)

::::::::: I agree that I do not have the age breakdown of that list - and though the info is there, I simply don't have enough free time to do that. So I will freely grant that until someone does that research, we don't exactly know the population size. But let's guess that it's 200 out of the near 500 we see listed - that's not unreasonable. But this is a list of mostly US and UK singers. Where are the comparable lists of Japanese, Croatian and Italian singers, who also no doubt have a tendency to succumb to [[Hollywood Values]]? How do we find out about the number of 'young hopefuls' in millions of towns around the world, rehearsing with a band in a garage somewhere, but not famous enough to be listed on Wikipedia, but very likely 'partying' and in the right age range? And that's just rock and pop singers - we haven't covered rappers, actresses or TV 'people' (I'm guessing chat show hosts, weather girls, news readers, etc might be just as susceptible to [[Hollywood Values]]?. IMDB lists well over half a million actresses - but again, only a portion of those will be in the right age range. But a very large number will be, as the age range is the functioning age range for most actresses. So while I'm openly admitting all these flaws in my own data, can you at least admit that your guess of 50 is very likely to be orders of magnitude off? That the final numbers could easily be in the thousands or even tens of thousands internationally? As you asked me - yes or no? Is that possible? [[User:EngelUmpocker|EngelUmpocker]] 22:01, 22 May 2008 (EDT)

:::::::::: EngelUmpocker, I have an open mind about this, and don't have any ax to grind. But I don't see the existence of thousands of touring female singers in their 20s. I just don't. The market wouldn't support it. There may be only 50. Or perhaps 100 or 200. But much more than that? I'm open to the possibility, but just don't think it's that many. Even many famous rock bands struggle to have successful tours, and most rock stars seem to be male.

::::::::::: Me too on the open mind - I know I don't have the data, but like you I have a 'hunch' about the numbers, and since our hunches are so different, I want to see if we can find a way to agree on the possible ''range'' of numbers if nothing else. Let's look at the phenomenon of 'Girl Groups' just for a moment - all girl pop groups. Spice Girls (5 members), Destiny's Child (3), Dream (4), Play (4), B*Witched (4), Eternal (3), En Vogue (4), SWV (3), Sugar jones (5), No Secrets (5), No Angels (4), Pussycat Dolls (5), Girls Aloud (5), Sugababes (3), Bardot (4), Eden's Crush (5) and The Cheetah Girls (3). A total of 69 young women, successful, all within the age range, and touring. If that's the case with just a few "Western' 'Girl Groups', can you imagine how many more young women around the world are in the business? [[User:EngelUmpocker|EngelUmpocker]] 14:09, 23 May 2008 (EDT)

So that confirms there are at least 50 popular female pop stars in the genres of rock, pop, and country and in the English speaking world (I'm not going to start looking for young pop stars in China, India, Japan, and other places that I'm sure have many such singers we'll never hear of). All of these, I'm pretty sure, are currently under 30. I don't know of any who have breast cancer, but I don't know that they don't. If we can identify a significant number of them who do, or identify others not on this list who maybe should be who do have breast cancer, we can have some numbers to work with. They will not be exact, and won't in themselves confirm any hypothesis, but they will be a good start, and give us something to work with. After all, the title of this is "mystery"; it doesn't claim to give definite answers and we shouldn't demand any at this stage

::The first is easy enough, we can limit the pool from performers born since May 1978 to born between May 1978 and May 1988. That will remove teenagers (though should any Hollywood teens turn out to have breast cancer they will have to be removed from the analysis as well, which could end up working against your hypothesis). Since the vast majority of those listed above are in their 20's the list won't change dramatically, I think.

::The second part is harder. Analysis of schedules and trying to make judgment calls on who is likely to hide a diagnosis is very subjective, particularly as we have no strong evidence of stars keeping it a secret. In fact, as a diagnosis of cancer is a surefire ticket to the cover of People magazine, they have an incentive to go public. Hollywood types thrive on publicity. Without compelling evidence of a statistically significant number of Hollywood women keeping cancer diagnoses secret, I don't see how it can be worked into even an informal study. If the results show a higher instance of breast cancer in this group, as you seem to think will be the case, then these potential secret cases won't matter, as we will already have established a higher rate without them. If they're lower for some reason, then that could be a sign that indeed some are not disclosing it. But I don't see any way of working theoretical non-disclosure into the study. Especially since if we make our study group too small, the probability of getting skewed results is greater. [[User:Cfred|Cfred]] 22:05, 20 May 2008 (EDT)

I'm not sure why the limit is to born after 1978 for the population. Minnie Riperton was born in 1947. The people should include individuals like Michelle Phillips, Carole King, Florence Ballard, Mary Wilson, Betty McGlown, Diana Ross, Barbara Martin, Cindy Birdsong, Jean Terrell, Lynda Laurence, Scherrie Payne, Susaye Greene, Martha Reeves, Rosalind Holmes, Annette Helton, Gloria Williams, betty Kelly, Lois Reeves, Sandra Tilley, etc... all the way through Victoria Beckham, Melanie Brown, Emma Bunton, Melanie Chisholm, Geri Hallwell. Why limit ourselves to the list of what is popular now - the list given has many more. There's a Colombian singer on the list too - so what about Anasol, Carolina la O, Fanny Lu, Carolina Herrera, Shakira, Toto la Momposina, Veronica Orozco. There's an Australian soap opera actress - so what the women listed in http://en.wikipedia.org/wiki/Category:Australian_soap_opera_actors though I'm not sure why one would limit to just soap operas - so http://en.wikipedia.org/wiki/Category:Australian_television_actors would be better to go from. There's an Australian pop singer in there too - so http://en.wikipedia.org/wiki/Category:Australian_pop_singers should be added too. Realize that this is just the list of what Wikipedia considers notable. I am sure that PJR can attest that there are many more than 83 female pop artists in Australia since the 1940s. --[[User:Rutm|Rutm]] 00:49, 21 May 2008 (EDT)

::: EngelUmpocker, your list of 69 might also be good. But the same analysis above applies. By the way, there is no way that there are a half million touring female singers. The entire New York City area, as large and entertainment-oriented as it is, can probably support only a handful on any given weekend.

::::Andy, the "half million" refers to IMDB listed '''actresses''', which is where I thought this article had originally begun. We have recently only been debating female singers - "let's now try to begin estimating how many of the half million actresses listed on IMDB are relevant", is what my point was? [[User:EngelUmpocker|EngelUmpocker]] 15:40, 24 May 2008 (EDT)

:::::: It most definitely would include people who would not be 'relevant', but it's a start. We must start to quantify these numbers ''somehow'', as otherwise it's just an assertion - and I know you don't want your theory based on an assertion. But over a half a million is a good start. Certainly I imagine there's tens of thousand of those who would fit our profile and are, or were professional actresses. Certainly, we can see that there's at the very least hundreds of touring young women in the rock and pop business in the 'Western' world. Next I hope to quantify jazz singers, classical musicians, circus performers and Broadway show-style performers. [[User:EngelUmpocker|EngelUmpocker]] 16:57, 24 May 2008 (EDT)

:As I said, yes, the list of all actresses has far more irrelevant names than relevant ones. But you're starting with a valid list of half a million - would you imagine even 1% of that list is valid? That's 5,500 individuals. And starting with the NYC phone book would be better than starting with a list displaying nothing but [[selection bias]], as you have here, since that would be essentially random, unlike your list. The key to understanding all statistical analysis is understanding that one's sample MUST be random, as countless others have pointed out above.

:Hello? [[User:Wandering|Wandering]] 23:10, 19 May 2008 (EDT)

::: I'd appreciate it if you answered my question, rather than make assumptions about my beliefs. [[User:Wandering|Wandering]] 20:19, 20 May 2008 (EDT)

:::: Do you plan to answer my question? [[User:Wandering|Wandering]] 20:00, 24 May 2008 (EDT)

:::::: I must therefore assume you do not have the answer. Oh well. [[User:Wandering|Wandering]] 20:29, 24 May 2008 (EDT)