Dr. Lockheart had hoped to remove the potential for tester bias by separating the surveying aspect of the study from the educational testing part
The testers, who do not know whether a child is homeless or not
meet the children at a school after the surveyors get the demographic information