I'm speaking of averages. On average, a two parent home is much better for children. Of course there are exceptions. Pointing out exceptions to robust statistical data is the most common fallacious statistical argument.
Statistical data don't tell why. Divorce could be the effect of the real problem (alcoholism or mental disorder) and staying married could actually have made things worse.
The thing is, once you got to decide, statistics say nothing useful, it's your situation or, if you prefer, your exception, your anecdote, what really matters.
