How did they contract and track poor people? If they tracked poor people using census data, or relied on poor people who have phones or addresses, then there is a large selection bias to this study: it totally overlooks those who are homeless, who lack phones, people without SS #'s, people who generally don't show up on census data and represent a larger problem than those who are poor "on the fringes."
Sorry for the extra post but there is so much ... uh ... well ... that you miss some the first time through.

DUDE, WHEN STUDYING PEOPLE WHO GOT JOBS, STAYED IN SCHOOL UNTIL GRADUATION FROM HIGH SCHOOL AND GOT MARRIED AND STAY MARRIED HOW MANY DO YOU THINK ARE HOMELESS OR DON'T HAVE SOCIAL SECURITY NUMBERS?

Out of all the people in poverty who did these three things, how many are likely to be in the group you note? MORE THAN 1 IN 200? LOL...dude, you are at least funny.