Original Research and Todd Akin Options

[Ottava]

Regardless of how you feel about Todd Akin, Republicans, or abortion, I think there is something else to focus on when it comes to Wikipedia. The Todd Akin rape and pregnancy controversy page has some wonderful original research but the kind that is rarely talked about.

"Related news articles cited a 1996 article in an obstetrics and gynecology journal, which found that 5% of women who were raped became pregnant, which equaled about 32,000 pregnancies each year in the US alone.[6] A separate 2003 article in the journal Human Nature estimated that rapes are twice as likely to result in pregnancies as consensual sex.[7]"

The sites used are blogs, and are used only to try and put forth studies.

It is easy to find articles on both sides quoting all sorts of statistics, especially this WSJ piece saying: "One article, written by Jack Willke, a Cincinnati physician and antiabortion campaigner and published in an antiabortion group's newsletter in 1999, concluded that "assault rape pregnancies are extremely rare" and that pregnancy could result from as few as one in 1,000 cases of rape. Dr. Willke said Monday that he stood by his article."

This is not the only one quoting a differing statement, but one of the most prominent (WSJ being prominent).


The reason why I call this original research is that you put a bias by quoting some loosely connected pieces to make a claim that is definite where the own pieces do not have that definite statement. The Politico article even states: 'But Gottschall did warn that methodological problems mean the numbers “aren’t carved in a stone.”' That is no where in the Wikipedia piece quoting the story.

The sources used are only tangentially connected, or do not have the same tone that Wikipedia has. Thus, you can put forth original research while having the appearance of citations. This happens quite often among science articles, and is common when people write a statement then try to add a source to it later to protect the statement.

[milowent]

The reason why I call this original research is that you put a bias by quoting some loosely connected pieces to make a claim that is definite where the own pieces do not have that definite statement.

This story is moving so fast that mainstream news articles are appearing almost as fast as wikipedia editors come up with this stuff, its one paid and one (hopefully) unpaid group googling the same sources and making the same connections.

I was able to whip up a bio on Fred Mecklenburg (author of a 1972 article that may be credited for perpetuating the no-juices-no preggers theory) pretty quickly, and the St. Louis Dispatch had posted an article basically connecting all the same dots.

http://en.wikipedia.org/wiki/Fred_Mecklenburg

[Ottava]

The reason why I call this original research is that you put a bias by quoting some loosely connected pieces to make a claim that is definite where the own pieces do not have that definite statement.

This story is moving so fast that mainstream news articles are appearing almost as fast as wikipedia editors come up with this stuff, its one paid and one (hopefully) unpaid group googling the same sources and making the same connections.

I was able to whip up a bio on Fred Mecklenburg (author of a 1972 article that may be credited for perpetuating the no-juices-no preggers theory) pretty quickly, and the St. Louis Dispatch had posted an article basically connecting all the same dots.

http://en.wikipedia.org/wiki/Fred_Mecklenburg

On one side, you have that study. On the other, you have attacks on that study, and the only way they can attack it is with this study:

http://www.ajog.org/article/S0002-9378%289...0141-2/abstract

It says 5% of rape victims not on contraceptives are impregnated, and also says there are 32000 rape pregnancies. That means there were 600,000+ rapes in the US for "fertile women" where reported rapes in the US reported were a total of 90,000 for all women (based on an FBI estimate, other groups estimate differently).

The ramifications of the study are astounding - at least 6 times the amount of rapes go unreported, and probably upwards of over 15 times when you figure in the amount of people who were on birth control, condom use by rapist, lack of ejaculation, etc. When you figure in the rest, the number of women raped is at least 2 million per year under their numbers. If you figure over 30 years of a person's life fitting the area they focused on, that would mean about an 80% chance a woman will be raped.

The numbers are really, really implausible, and there is nothing to suggest that their math in any regard was correct. That is the unfortunate thing about "studies" that are more surveys of a group of people without a control group, without any real diversity, etc., and scientists more eager to make claims than to try and see if the claims make sense.

Yet there are doctors on tv using the study to say that rape pregnancies are actually more common than normal pregnancies. They don't say that it would only be 5% if there was no birth control used, if the guy fully ejaculated, etc. (the number drops to well below .1%).


If you want to see how the numbers are even more startling - they start that over 15,000 of the rape pregnancies result in an abortion. Abortions after rape in the US account for less than 1%. Using the 15,000 number, that would mean there are more than 1.5 million abortions in the US per year. There were 3 million babies born last year. That would mean that one in every three pregnancies would result in an abortion. That is higher than many estimates I've seen, but there are no actual statistics on abortions because many states ban taking them.

So the numbers are all over the place. I've heard that people on both sides are afraid to find out what the actual amount of abortions are, which is necessary to determine any of these numbers. But Wikipedia doesn't care - it turns vaguery and uncertainty into fact as long as they have a blog willing to push a claim without care.

This post has been edited by Ottava: