Talk:Mars effect

From Wikipedia, the free encyclopedia

This is a controversial topic, which may be disputed. Please read this talk page discussion before making substantial changes. (This message should only be placed on talk pages.)

The purpose of doing a statistical analysis is to draw conclusions based on the generally-accepted levels of probability that allow one to state (on the basis of those accepted levels of probability) whether something is true or false. Between the extremes, there is an area where the analysis is inconclusive.

Drawing "conclusions" is not a "purpose" of statistical analysis. There is also no such thing as "generally-accepted levels of probability". One can speak in terms of standard deviations; data that is more than two standard deviations from the expected value tends to be considered meaningful, but it can still be generated by random processes. Any statistical results that fall close to the expected results are "inconclusive" because they don't imply anything other than that the data appears to be acting normally. Perhaps some other experiment might produce different results, but we have no way of knowing that. ☮ Eclecticology 02:04, 2004 Jan 2 (UTC)

I didn't want to get too technical, but the 'accepted levels of probability' I was thinking of were confidence intervals. But we're really just having a semantic argument about the meaning of conclusion here. My SOED defines the meaning I'm thinking of as "A judgement or statement arrived at by reasoning; an inference, a deduction." You seem to be using it in some stricter sense, tantamount to 'proof'. For example, if I had a sample that had a 95% probability of being produced by chance, I would 'conclude' that the sample was produced by chance; one that had a 5% probability of being produced by chance, I would 'conclude' was not produced by chance; and one that had a 50% probability of being produced by chance, I would regard as 'inconclusive'. That's why I see the purpose of statistical analysis as being to draw conclusions - otherwise it's like having a legal trial but not pronouncing a verdict ... just pointless. Spellbinder 08:56, 2 Jan 2004 (UTC)

I did accept that "conclusion" was better than "proof", but perhaps neither is the appropriate word. Perhaps "inference" might be better, but statistical inference is qualitatively different from logical inference. I suppose we could be said to be having a semantic argument, but I would tend more to categorize it as an argument on the quicksand nature of statistics. We really can't conclude anything from statistics. I do accept that a 50% confidence interval exists at ±1sd from the norm and a 95% confidence interval at ±2sd from the norm. "Conclude" implies a certain finality. "Probability of being produced by chance" strikes me as backward; one can argue in many contexts that all events are produced by chance. And that possibility can be defined as the sum of all improbabilities. That gets us nowhere. (I hope that my current reading of The Hitchhiker's Guide to the Galaxy in not too much of an influence here.:-)) The 95% confidence interval suggests a meningful linkage to the norm rather than a merely chance event. Such results are important because they are "inconclusive", or "inconsequential" or even "meaningless". Legal trials are indeed driven by a compulsion to have a verdict, and an unwillingness to accept enigmatic results; to get there courts often must rely on simplistic versions of the truth. In reference to UFOs I am content to accept the letter "U" in the term as implying that many of them will remain just that: unidentified. I can accept that for most astrological phenomena the scientific proof does not exist, but that only leaves them unproven rather than false. That uncertainty or absence of conclusions does not bother me. ☮ Eclecticology 00:45, 2004 Jan 3 (UTC)

To say that a statistical analysis is anomalous (i.e. irregular, abnormal) is not meaningful. It may be that the results are anomalous, but that implies the issue has been pre-judged and that those results do not fit one's preconconceptions of what the results ought to be.

I can agree that "anomalous results" would be preferable to "anomalous analysis", but "pre-judged" is an improper inference. Any statistical results that deviate significantly from the statistically expected value are anomalous. The superficial preconception here would be that 25% of the Mars positions in a chart should be in each quadrant. If in an experiment noticeably more or less than 25% of the Mars observation were in that fourth quadrant, that would be an anomalous result. On the basis of a single experiment, one could say nothing more than that further investigation is warranted. ☮ Eclecticology 02:04, 2004 Jan 2 (UTC)

I can see your point, but I still think 'inconclusive' is a much better word - it encapsulates why the issue is so contentious i.e. that the two sides (astrologers and sceptics) are drawing, or at least saying the door is left open to, the opposite conclusions. Also, 'inconclusive' is non-POV, whereas 'anomalous' (with its connotations of 'abnormality') has a hint of the sceptical POV. Spellbinder 08:56, 2 Jan 2004 (UTC)

The concept of proof is not relevant here - statistics can never prove anything; nor can they disprove anything.

I can accept that, and use the word "conclusion" instead of "proof". Apart from that I stand by my changes. When data does not allow one to draw the conclusion that a statement is true, one cannot draw the conclusion that the statement is false. In logic I believe that that is the fallacy of the excluded middle. ☮ Eclecticology 02:04, 2004 Jan 2 (UTC)

"Correlation is not causality" - very true, but the paragraph changed is not about causality; it´s about Gauquelin's testing whether the correlations predicted by astrology are statistically verifiable. Spellbinder 08:34, 1 Jan 2004 (UTC)

The Mars effect pointed to an anomalous correlation of the Mars position is the chart of athletes. That is a correlative statement. What I removed was a statement that this "influenced human destiny". That is a causal statement. I believe that it is a fair statement to say that astrologers today tend more toward an acausal vision of their subject. ☮ Eclecticology 02:04, 2004 Jan 2 (UTC)

Well, I don't agree that "influenced human destiny" is a causal statement. I definitely do agree that astrologers don't give any causes as to how the positions of heavenly bodies , well, influence human destiny ... but by saying there's a correlation, surely they're saying there's an influence, even if they don't know the cause. But perhaps we're just down to semantic argument again. But what's really important here is that the sentence about the Mars Effect being true means that a heavenly body can influence human destiny is meant to emphasize to the normal encyclopedia readers just why the Mars Effect has been so contentious and fought over so much. If you want to take it out, perhaps you can think of a way to emphasize this using your own words. Spellbinder 08:56, 2 Jan 2004 (UTC)

Hmm, I see now that influence is a technical term in astrology, so I agree that the phrasing of "influenced human destiny" is unacceptable. The latter point about emphasizing why the issue is so important and contentious still applies though. Spellbinder 09:35, 2 Jan 2004 (UTC)

Contents 1 NPOV 2 Incomplete 3 Arbitrary definition of "eminence" 4 Added "Cleanup-Confusing" tag

[edit] NPOV

The tone of this article isn't proper for an encylopedia and reflects the author's opinion on the issue. A prime example is: "Read it a second, third, or fourth time, though, and you will see that the initial appearances are deceptive." Not appropriate. Subtler comments like "...all of which, in retrospect, could have been avoided." still reflect a certain point of view. --Tothebarricades.tk 00:57, May 16, 2005 (UTC)

It needs to be more encyclopedia in tone. Bubba73 (talk), 01:18, 1 December 2005 (UTC)

I agree, the article should say what needs to be said, not argue one way or another. GregorB 10:28, 4 March 2006 (UTC)

I removed the "all of which, in retrospect..." bit, and reworded your first objection. I agree the article still seems a bit NPOV, heavily biased towards the skeptical side. I'd like to see more of Gauquelin's take on the theory. I read one of his books a couple of decades ago, but don't remember much about it, except that he studied much more than the "mars effect" (finding, for instance, similar effects with other astrological bodies and professions, such as Jupiter among politicians and the Moon among writers.) Aaronwinborn 16:04, 16 March 2006 (UTC)

This article still has serious NPOV issues and is essentially a pro-Mars effect article. It even makes statements outright claiming what is right and wrong. I'm putting up an NPOV tag. -Nathan J. Yoder 06:02, 22 March 2007 (UTC)

[edit] Incomplete

No mention of Ertel? No mention of the French Study? This article isn't POV, it's simply incomplete. Mystylplx 04:35, 31 March 2006 (UTC)

Also, Geoffrey Dean's conclusion that the Mars effect is a social artifact, caused by the faking of birth data by hopeful parents, should be mentioned. Without it, this is definitely POV. 85.144.128.34 11:11, 29 April 2006 (UTC)

Not just incomplete but completely unbalanced by having a 'Criticism' section twice as long as the 'Introduction', and there is not even a main body. Needs a lot of work. Aquirata 07:25, 28 June 2006 (UTC)

[edit] Arbitrary definition of "eminence"

In this page, I miss the criticism that Gauquelin's definition of eminence is very flexible. If you are allowed to draw the line between eminence and non-eminence within a group of sportsmen at an arbitrary place, you can easily get "evidence" for an eminence effect by choosing the optimum place for the division. Kurtz, Nienhuys, and others say that is what Gauquelin did and Ertel still does. Thus, basketball players turned out to be not eminent at all whatever their achievements Example source: Nienhuys posting. That is the most famous example, but there are other examples of subgroups with arbitrary borders that can add up to such an effect, such as 268 Italian aviators (which are all excellent) and Italian soccer players (which need to have played in an international game to be famous) Example source: Dobyns posting. I never heard of any attempt to refute this. --Hob Gadling 13:00, 25 July 2006 (UTC)

I agree that it should be in there. Do you want to put it in? Bubba73 (talk), 15:50, 25 July 2006 (UTC)

Done. I also structured the criticism section, some of which is criticism of criticism of criticism. --Hob Gadling 12:25, 17 August 2006 (UTC)

There is a reference to a newsgroup posting by Nienhuys in the article. Does the article by the same person (in the external links) also say that? If so, it would be a better reference than a newsgroup posting. Bubba73 (talk), 22:46, 19 August 2006 (UTC)

[edit] Added "Cleanup-Confusing" tag

This page needs a lot of work; it doesn't make much sense, and I know what it's supposed to be talking about. It also doesn't include the most recent commentary, especially Geoffrey Dean's conclusion that the Mars effect was a result of parents lying about their children's birthdays. I don't feel up to the task of changing it myself, though. - CronoDAS 03:19, 18 August 2006 (UTC)