Results of a functional analysis indicated that the hand mouthing of a woma
n with developmental disabilities was maintained by multiple sources of con
trol (sensory stimulation and access to a leisure item). Further assessment
revealed that access to several other items also produced high rates of ha
nd mouthing. However, direct observation conducted in the woman's home indi
cated that none of these items was delivered contingent upon hand mouthing.
When the consequence observed most frequently in the home was incorporated
into the functional analysis, rates of hand mouthing were no higher than t
hey were during an alone condition. We concluded that hand mouthing, althou
gh maintained by automatic reinforcement, was also susceptible to social co
ntingencies when exposed to them during assessment, thereby producing a par
tially false-positive outcome.