There many problems of methodology inherent in the 2005 AARC study. Here are some:
Distortions in Self-Report Data
-Social Desirability Bias
"Subjects who are influenced by this bias work overtime trying to create a favorable impression(DeMaio, 1984)"
Is it socially desirable to admit that you're using illegal drugs and drinking after your family dropped $50 000 on treatment for you? How about after leaving a facility that maintains links to your family for years after your stay there is completed?
-Response Sets
"Some people tend to agree with nearly everything on a questionnaire(Krosnick and Fabrigar, 1998)"
Experimenter Bias
"Experimenter bias occurs when a researcher's expectations or preferences about the income of a study influence the results obtained."
The experimenters were the Executive Director of AARC; another staffer who is also a former client, wife of a former client, and sibling of multiple former clients; a board member of AARC who is also the parent of a former client turned staffer; a paid consultant to AARC; yet another faculty member of the school that has traditionally provided endorsements for AARC, as well as the Executive Director's PhD, and has in turn received money from AARC.
No double-blind was used to reduce experimenter bias.