Validation problems and technological change in an input‐output model

Thornton, Barry; Sorli, R. M.

doi:10.1080/0020739820130601

Cited by 1 publication

(1 citation statement)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One of the significant issues involved in input/output modeling is how to deal with missing information from surveys. Thornton and Sorli (1982) discussed problems associated 5.16 with validating the input/output model after forecasted data had been substituted for missing information. According to Mickle (1998) the most important point is predicting past outcomes with the model.…”

Section: Verification Validation and Calibrationmentioning

confidence: 99%

Model and Analytic Processes for Export License Assessments

Thompson¹,

Whitney²,

Weimar³

et al. 2011

View full text Add to dashboard Cite

This report represents the U.S. Department of Energy Office of Nonproliferation Research and Development (NA-22) Simulations, Algorithms and Modeling (SAM) Program's first effort to identify and frame analytical methods and tools to aid export control professionals in effectively predicting proliferation intent-a complex, multi-step and multi-agency process. The report focuses on analytical modeling methodologies that alone, or combined, may improve the proliferation export control license approval process. It is a follow-up to an earlier paper describing information sources and environments related to international nuclear technology transfer. This report describes the decision criteria used to evaluate modeling techniques and tools to determine which approaches will be investigated during the final 2 years of the project. The report also details the motivation for why new modeling techniques and tools are needed. The analytical modeling methodologies will enable analysts to evaluate the information environment for relevance to detecting proliferation intent, with specific focus on assessing risks associated with transferring dual-use technologies. Dual-use technologies can be used in both weapons and commercial enterprises. A decision-framework was developed to evaluate which of the different analytical modeling methodologies would be most appropriate conditional on the uniqueness of the approach, data availability, laboratory capabilities, relevance to NA-22 and Office of Arms Control and Nonproliferation (NA-24) research needs and the impact if successful. Modeling methodologies were divided into whether they could help micro-level assessments (e.g., help improve individual license assessments) or macro-level assessment. Macro-level assessment focuses on suppliers, technology, consumers, economies, and proliferation context. Macro-level assessment technologies scored higher in the area of uniqueness because less work has been done at the macro level. An approach to developing testable hypotheses for the macro-level assessment methodologies is provided. The outcome of this works suggests that we should develop a Bayes Net for micro-level analysis and continue to focus on Bayes Net, System Dynamics and Economic Input/Output models for assessing macro-level problems. Simultaneously, we need to develop metrics for assessing intent in export control, including the risks and consequences associated with all aspects of export control. v Abbreviations and Acronyms ACE agent-based computational economy ACH Analysis of Competing Hypotheses BIS Bureau of Industry and Security CA Content Analysis CGE computable general equilibrium

show abstract

Section: Verification Validation and Calibrationmentioning

confidence: 99%