Tuesday, July 29, 2014

The Problem with Statistics

Dear Book Author:
I read your book.  Thanks.  It was interesting.

I work with numbers a lot.  That said, I have discovered that there are many ways to go wrong with numbers.  I know, having gone wrong, and seen others go wrong, a lot.

I know you spent over 8 years on this issue, and I have spent the last month trying to figure out the nuances of the US divorce statistics.  As you acknowledge, this is not as straight-forward as it might seem.  Every married couple has potential for divorce as long as both of them are alive.  I agree that the common divorce statistic or 50% of all marriages end in divorce, is mostly speculation.  

I understand that a much lower divorce rate would be great news about marriage. I have also published a book, and as soon as it came out, I began to worry about mistakes I might have made.  I have marked one of my copies with numerous things I will change in the next edition.  Likewise, I understand that you have a vested interest in defending your findings against all arguments.

 In short, I wish you were right, but it sure doesn't look that way to me.  You claim, understandably, that the Census Bureau shows that 72% of married folk are still married to the same spouse, suggesting a divorce rate of less than 28%.  You claim it must be considerably less than 28% due to the marriages ended in death of a spouse, and this sounds good to me, if I could believe it.

I know that in the back of the book you address some obvious complications.  But for some of these, you ignore their ramifications.  Here are some things you did NOT seem to take into account on marriage-divorce calculations:
  1. Cohabitation.  People who co-habitate never divorce, because they never had a legal marriage, and in most cases didn't marry for that very reason-- no marriage= no divorce.  You argue that co-habitants eventually get married.  Perhaps they do marry, but they don't necessarily marry the person with whom they cohabited.  What is your basis for ignoring these?
  2. Annulments.  Whereas, again these technically aren't divorces, they are the equivalent.  The numbers aren't large, but deserve to be counted.
  3. Multiple marriages.  A person marrying 3 times has at least two divorces (unless they are Mormon- joke).  Yes but it's one person.  Remember, we are counting divorces, not people.
  4. Separations (long term ones).  Again, these aren't divorces, but they aren't exactly "good news" for marriages.
So, is the divorce rate 50%?  No, I don't think so, but it is likely to be closer to 50% than to 25%.  Does anyone know the real number?  Not as far as I can tell.  

Again, as an author, I don’t expect you to do anything but defend your stats.  As a Christian I feel I need to let you know of my objections, and ask you to perhaps caveat your views a bit.

I have seen one principle that we Christians ought to remember.  Countering bad information with good information is useful; countering bad information with equally bad information is worse than leaving it alone.

Divorce is a terrible thing.  Christians ought to do all we can to counter it, first inside the Church, but also outside it.  "I hate divorce."  G^d tells us (Mal 2:16).  We ought to do likewise.


No comments:

Post a Comment