[Skip to Navigation]
July 1992

Methodological Issues in the Assessment of Human Immunodeficiency Virus—Related Cognitive Impairment-Reply

Author Affiliations

Neurological Institute 710 W 168th St New York, NY 10032
New York, NY

Arch Gen Psychiatry. 1992;49(7):587-588. doi:10.1001/archpsyc.1992.01820070081018

In Reply. —  We welcome the opportu- nity to clarify the issues raised by Miller et al.Several questions were raised about the possible influence of other demographic features on neuropsychological test performance. All participants in our study were fluent English speakers, as demonstrated by their ability to fully comprehend informed consent; participate effectively in extensive psychiatric, psychosocial, and psychosexual interviews; and undergo neuropsychological testing. We still were concerned about the distribution of subjects for whom English was not a first language and described it fully in the article. This variable was used as a main effect in multivariate analysis of variance and regression analyses. Despite small numbers and reduced power, significant first language effects were noted. The comparability of measurement error is an issue that is not specific to language and can be questioned in any analysis. We believe that the treatment of the language issue was sufficient in the

Add or change institution