The Federal Government bases the working poor level as anything lower than an absolute number. The researchers chose to base their study on a relative number that is derived from a ranking of real world assumption. They experimented with many different definitions of what represents the working poor but ultimately use the highest ranking definition, since it according to them, reflects the most accurate, real world scenario.

More than anything, the study is meant to convince those in decision making positions, that a relative number is more reliable than the absolute number used by the Federal Government. The relative number works well because the assumptions used to arrive at it help guide the decision making by pin pointing causes.

Sadly, I can't access the actually study to read it. I assume the study addresses the claim that most of America's poor actually work. From the Science article and news release there is only a discussion of how to accurately define a poverty level among those who do work (defined as more than 17 hours a week with other adjustments). I don't follow how that shows that most poor are actually working. I would also like to see an analysis of motivations and mobility. How many of these working poor are that way for more than a year or two and how many actually try to move out of that classification (through education and career training).