The results are refutable if the initial suppositions are called to question. Did they randomly select 13 vagrants? No they selected hardened street people... probably only males.
Taint the test sample with some teenaged girl runaways turning tricks for a bag of H, and you might get different results.
There are more glaring weaknesses in the effort than just the randomness of the selection of the vagrants.
Thirteen is hardly a very large population, especially given the ambiguous result that 11 were "off the street". Some of the original 13 were in "hostels" and some in "shelters". That's hardly a description of lifting someone out of poverty.
For all we know, the results would have been identical even without the cash distributions. That brings up a second glaring weakness in the so-called "study". Where was the "control group" which was monitored to demonstrate that the lack of money would have resulted in continued homelessness? There is no data at all to demonstrate that the efforts of those making the study had any effect whatever. For all we know, the group might have been better off without the money.
What passes for "science" these days provides a full explanation for why we are suffering the global-warming hoax.