Matches in Harvard for { <http://id.lib.harvard.edu/aleph/006114864/catalog> ?p ?o. }
Showing items 1 to 19 of
19
with 100 items per page.
- catalog abstract "This paper examines the problem of formulating measurement strategies for and interpreting measurements of dependent and independent variables in complex multidimensional systems. In particular, the role of a nonseparable relationship between the dependent variable in a causal relationship and the multiple independent variables is discussed. It is shown that there exists a set of nonseparable functional relationships between the dependent and independent variables which pose a serious challenge to the relevance and robustness of relationships inferred from localized measurements. The risk of treating nonseparable systems as if they were separable is examined. A useful conceptual model for the measurement of complex systems is a sampling process. The cost in sampling efficiency imposed by multidimensional systems is considered, and the benefit of prior insight into the system's behavior is incorporated into the formulation of the sampling process. The coupling between the observer's prior analytic assumptions about the system under study and the measurement process can thus be quantified, through the formulation of the multidimensional sampling process. This approach is contrasted with the incorporation of prior insight into the measurement process using Bayesian statistics.".
- catalog contributor b8562393.
- catalog contributor b8562394.
- catalog contributor b8562395.
- catalog created "1994.".
- catalog date "1994".
- catalog date "1994.".
- catalog dateCopyrighted "1994.".
- catalog description "Includes bibliographical references (p. 25).".
- catalog description "This paper examines the problem of formulating measurement strategies for and interpreting measurements of dependent and independent variables in complex multidimensional systems. In particular, the role of a nonseparable relationship between the dependent variable in a causal relationship and the multiple independent variables is discussed. It is shown that there exists a set of nonseparable functional relationships between the dependent and independent variables which pose a serious challenge to the relevance and robustness of relationships inferred from localized measurements. The risk of treating nonseparable systems as if they were separable is examined. A useful conceptual model for the measurement of complex systems is a sampling process. The cost in sampling efficiency imposed by multidimensional systems is considered, and the benefit of prior insight into the system's behavior is incorporated into the formulation of the sampling process. The coupling between the observer's prior analytic assumptions about the system under study and the measurement process can thus be quantified, through the formulation of the multidimensional sampling process. This approach is contrasted with the incorporation of prior insight into the measurement process using Bayesian statistics.".
- catalog extent "30, [12] p. :".
- catalog isPartOf "Working paper (Harvard University. Graduate School of Business Administration. Division of Research) ; HBS 94-004.".
- catalog isPartOf "Working paper / Division of Research, Harvard Business School ; 94-004".
- catalog issued "1994".
- catalog issued "1994.".
- catalog language "eng".
- catalog publisher "[Boston] : Division of Research, Harvard Business School,".
- catalog title "Nonseparability, sampling, and the observation of multidimensional systems / by M.C. Moldoveanu, H.H. Stevenson, J.W. Pratt.".
- catalog type "text".