Monday, May 24, 2010

Once in a Lifetime data gathering techique is totally DUBIOUS

People who follow the $1 billion Domestic Violence industry may notice that even in India  women’s organizations have started to  present Domestic Violence statistics in Once in a lifetime statistical method. Below are some links that highlight this method in news reports .

a) News report Link

The above report says that 30% of women (or to make it sound more serious 1 in 3 ) face domestic voilence "once in their lifetime" . To the common untrained mind this method may sound very credible and normal, but is it really so ? Lets look at the example below to find out. 

 The Claim
 Let’s start with an oft quoted statistic: - 40% of women have been suffered Domestic violence at least once in their lifetime. First  you see the time line has been extended from 2 years to 5 years to 10 years and now the whole lifetime . So how does this whole lifetime timescale benefit the feminists?
The Analysis
1) This means a woman ( Lets call her Savithri )  who has encountered DV in 1980 will also appear in this data , although she might not have encountered any violence for the past 30 years . That means if we plot the occurrence of DV on the Y axis and Timeline on the X axis then there will be only one occurrence of DV on the graph 30 years ago. So this method of representing data in order to give a false sense of frequency of occurrence increases the time scale enormously which is wrong .

2) Now Savithri's data will keep appearing in this DV victim list until Savithri dies in say 2040 at the age of 80 . So the DV data will still look high since until Savithri dies her data will remain valid. Savithri's data will only disappear when she is dead and she has no way to report that she has harassed once in her lifetime back. So the one data is kept alive from the time of occurrence of the act to the time the person dies. This in no way gives a true representation of the seriousness of the problem , in turn it tries to create false and bloated numbers which are essentially false. A person might have been afflicted by smallpox in 1960 , however small pox is now eradicated from the planet. Can we say at least 40% of people alive today faced smallpox at least once in their lifetime?

3) Lets assume for one second that Savithri was really beaten. But Savithri was beaten in 1980 it has been 30 years and there has been innumerable laws and legislation's passed against DV and harassment. However let us assume that in 2005 we have passed a DV law which was like a silver bullet and eliminated all DV in India. Even though today in 2010 DV is completely eliminated an there is no real need of the DV law , Savithri's Data from 1980 is used to show DV statistics . So until Savithri's death in 2040 it will shown that there are women who have faced DV( once in their lifetime) and hence there is need for the DV law . Speaking in different terms this means that even though the smallpox vaccination eradicated the disease from the face of the planet , the data gathering methodology ensures that it still shows up for a very long time . The effect of the remedy of the solution is very effectively eclipsed  due to the data gathering technique used .

4) So you see how the argument of having faced DV once in a lifetime is used to show data that is not relevant , but for the non-mathematical common man  this looks like a significant figure .

Did Savithri tell the truth ?
The question also remains that did Savitri really tell the truth about Domestic Violence? With over 98% of all 498a and about 80% of all DV cases turning out to be false can we really even believe Savithri when she claimed that she faced Domestic Violence in 1980 .

Now it is left to the intelligence of the readers , how the data gathering techniques are used to essentially fudge numbers in order to attract funds for Domestic Violence . In order to get a true representation of the problem DV stats must be gathered year wise or month wise and only based on convictions and not complaints.

No comments:

Post a Comment