| Autor: |
Rubin, Donald B. |
| Předmět: |
|
| Zdroj: |
Statistical Science; Aug2006, Vol. 21 Issue 3, p319-321, 3p |
| Abstrakt: |
The author reflects on the comments made by statisticians Edward L. Korn, Paul R. Rosenbaum and Stephen E. Fienberg on his investigation into censoring by death and the quality of life (QOL) of patients after a treatment using causal inference. He agrees on Korn's argument that distributional assumptions can be avoided if the principal strata are observed. The rank-sum test advocated by Rosenbaum avoids selecting a single value for the QOL. He counters the use of formality in conveying ideas supported by Fienberg. |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
