SUSSER, E.(http://cait.cpmc.columbia.edu/research/faculty/facu0330.html)
WITTKOWSKI, K.M. (1996-10-07) Division of HIV/AIDS (E-48), Centers for Disease Control and Prevention (CDC)
When analyzing ordinal data, we are faced with two problems: (1) Neither principal component analysis (PCA) nor sums of scores (SOS) are invariant against scale transformations. Moreover, PCA is inconsitent in case of competing risks and SOS requires knowledge of the relative weight and the correlation of the variables . (2) Classical rank tests (e.g. the sign-test and GEHAN's test, both asymptotic and exact) are inappropriate for data with varying precision (e.g. discretized or censored). They may have low power or exceed the level, depending on the distribution of precision.
A new approach is presented, where (1) rank scores are computed by a scale invariant general procedure generated by the semi-ordering of the (multivariate or censored) data and (where available) the semi-ordering of external calibration varibles these rank scores are weighted by non-parametric 'Fisher information' -type weights, and (3) different strategies are used to estimate the variance depending on the nature of 'ties' . The new tests can be seen as special cases of a single method that includes many well-known tests. Some terms commonly used are redefined and implications of this new classification of knowledge on the design of user-interfaces are discussed.
Applications based on recent studies include risk analysis (exposure to Schistosomiasis, infection with HIV), knowledge, attitute, and behavior evaluation (risk of HIV infection), survival analysis (e.g. duration of side effects), assessment of side effects (CMV treatment), quality of live (e.g. among patients at risk to develop Alzheimer disease). The author will spend the winter semester 96/97 at the Div. HIV/AIDS (E-48), CDC.
DESVARIEUX, M., WITTKOWSKI, K.M., DEVINCENZI, I., SUSSER, E.(internet-link: "http://cait.cpmc.columbia.edu/research/faculty/facu0330.html") (1996) A New Objective Method for Computing Risk Indices Applied to Cohorts of HIV-Discordant Heterosexual Couples . XI International Conference on AIDS(internet-link: "http://www.interchg.ubc.ca/aids11/homepage.html") Abstracts, Vancouver, CA, 1 :159 (Mo.C.1608)
EINSELE, H., EHNINGER, G., HEBART, H., WITTKOWSKI, K.M., SCHULER, U., JAHN, G., MACKES, P., HERTER, M., KLINGEBIEL, T., LÖFFLER, J., WAGNER, S., MÜLLER, C.A. (1995) Polymerase chain reaction monitoring reduces the incidence of cytomegalovirus disease and the duration and side effects of antiviral therapy after bone marrow transplantation. Blood(internet-link: "http://www.bloodjournal.org/") 86
SUSSER, E., DESVARIEUX, M., WITTKOWSKI, K.M. (199?) Reporting sexual risk behavior for HIV: a practical risk index and a method for improving risk indices. Am J Publ Health (in press)
WITTKOWSKI, K.M. (1988) Friedman-type statistics and consistent multiple comparisons for unbalanced designs. Journal of the American Statistical Association 83
WITTKOWSKI, K.M. (1989) An asymptotic UMP sign test for discretized data. The Statistician 38
WITTKOWSKI, K.M. (1992) An extension to Wittkowski. Journal of the American Statistical Association 87
WITTKOWSKI, K.M., DAMIANI, K., DBOUK, H., GHOLMIEH, G., HASHEM, A., HIJAZI, R., HUSSEINI, S., TABBAL, J.E. (199?) Knowledge of Risks of HIV Infection and Methods of Prevention among Women in the Greater Beirut Area. (submitted)
WITTKOWSKI, K.M., HUSEIN, M.H., TALAAT, M., BARAKAT, R. (199?) A new procedure to access individual risk of exposure to cercariae from multivariate questionnaire data. (in preparation)
Back to Homepage
SUSSER, E.(http://cait.cpmc.columbia.edu/research/faculty/facu0330.html)
XI International Conference on AIDS(http://www.interchg.ubc.ca/aids11/homepage.html)
Blood(http://www.bloodjournal.org/)