ENBIS: European Network for Business and Industrial Statistics

My abstracts

The following abstracts have been accepted for this event:

• Graphical Selection of Effects in General Factorials
  
  Authors: Pat Whitcomb Gary W. Oehlert
  Affiliation: Stat-Ease, Inc. & School of Statistics, University of Minnesota
  Primary area of focus / application:
  

  Submitted at 3-Apr-2008 13:18 by Pat Whitcomb
  Accepted (view paper)
  

  22-Sep-2008 15:20 Graphical Selection of Effects in General Factorials
  Many design of experiment practitioners are familiar with the half-normal plot introduced by Cuthbert Daniel (1959) to select model terms in the standard two-level factorial (2^k-p series) designs. We show an extension to the Daniel plot of effects that permits factors with more than two levels. Advantages of half-normal plot are: it is less intimidating than numerical methods, it is easy to understand and it provides a degree of protection against multiple testing.
• Reliability of the Visual Analog Scale as a Measurement Method of Economic Expectations
  
  Authors: Anna Stangl
  Affiliation: Ifo Institute for Economic Research
  Primary area of focus / application:
  

  Submitted at 7-Apr-2008 14:52 by Anna Stangl
  Accepted
  

  22-Sep-2008 14:00 Reliability of the Visual Analog Scale as a Measurement Method of Economic Expectations
  Economic expectations are playing a central role in contemporary economic theory as determinants of economic activity of agents. The measurement of economic expectations continues to be of particular interest to economists and statisticians in various fields.
  The paper introduces a new measurement method of economic expectations – the Visual Analog Scale (VAS) – and compares its reliability to the traditional three-category scale. Visual analog scale has not yet been applied to the measurement of economic expectations, although it is an attractive alternative to coarse category-rating scales and to the elicitation of quantitative responses. On VAS respondents can express their expectations on a continuum, which allows for a subtle distinction of their preferences.
  In the literature on economic expectations and their measurement, the main focus has been placed on validity criteria, particularly rationality and forecasting properties of economic expectations. The key research issue has been to link qualitative expectations data to the real quantitative outcomes. This paper adds to the literature of the measurement of economic expectations a comprehensive analysis of scale reliability. Four classes of reliability are examined: Parallel-forms reliability, test-retest reliability, internal consistency and inter-rater reliability.
  There are further arguments making this study interesting: (1) Previously VAS was merely applied in personal interviews in medical settings. In the present study VAS is applied in a sizeable self-administered Web-survey in an economic tendency survey. (2) The study tests an entirely new measurement method of economic expectations, which so far are measured qualitatively (with category-rating scales), quantitatively (by eliciting point estimates) or by subjective probabilities. (3) The data of the study are based on a real panel survey of economists, containing 2,470 observations from eight consecutive quarterly survey waves in the years 2005-2007, which makes possible to observe long-time effects in VAS response behavior.
  VAS was found to be a reliable measurement method of economic expectations and in several reliability classes even better than the three-category scale that is traditionally used in business surveys.
• Comparison between two methods of metamodel building (spline and kriging) for the efficiency calculation of a micro electronic component.
  
  Authors: Durrande De Crecy
  Primary area of focus / application:
  

  Submitted at 9-Apr-2008 14:07 by Nicolas Durrande
  Accepted
  

  23-Sep-2008 15:00 Comparison between two methods of metamodel building (spline and kriging) for the efficiency calculation of a micro electronic component.
  Production hazards usually lead to a difference between the theoretical dimension of an electronic component and the actual dimension observed after production. In most cases these differences aren’t problematic since all components produced comply with standards. However, in some cases, a percentage of the production doesn’t comply with a particular criterion and has to be disposed of. This allows us to define the efficiency of the production as the percentage of valid components at the end of the production process.
  
  In our study, we know the probability law of the thickness of the different coats making up the component. Furthermore we have a numerical simulator that gives certain characteristics of the spectral response of a filter according to the thickness of 10 coats. Unfortunately the evaluation of the response by the simulator is costly and the efficiency of the production process can’t be calculated with the simulator by a Monte-Carlo method. The solution is to build a metamodel based on a limited number of evaluations by the simulator.
  
  There are many ways to choose the points to be calculated by the simulator and lots of ways to build the metamodel. We study three kinds of space filling designs (Halton sequence, LHS with the criteria maximin and a classic design with arbitrarily chosen points) and three kinds of metamodels (spline type thin plate, ordinary kriging and universal kriging). Eventually the models are compared and two methods for the efficiency’s calculation are studied, one of these using the kriging variance.
• Bayesian Hierarchical Model for Integrating Multi-resolution Metrology Data
  
  Authors: Heidi Xia, Yu Ding, and Bani Mallick
  Affiliation: Texas A&M
  Primary area of focus / application:
  

  Submitted at 13-Apr-2008 17:33 by Yu Ding
  Accepted
  

  23-Sep-2008 12:20 Bayesian Hierarchical Model for Integrating Multi-resolution Metrology Data
  This article presents a Bayesian hierarchical model to integrate multi-resolution metrology data for inspecting geometry quality of manufactured parts. Our treatment includes using a Gaussian process for modeling metrology data at a single resolution level and establishing a linkage model to connect datasets at different resolutions. A unique problem in this metrology application is the misalignment between the low- and high-resolution datasets.
  To address this issue, an optimization procedure is first employed to roughly align datasets of different resolutions and then a neighborhood linkage model is introduced to link a high-resolution data point with a group of low-resolution data points. Improvements are demonstrated using the datasets from a milled sine-wave part, measured by two coordinate measuring machines of different resolutions, respectively.
• Migration in Structural Credit Rating Models
  
  Authors: Dr P. Avramidis
  Primary area of focus / application:
  

  Submitted at 18-Apr-2008 09:51 by Panagiotis Avramidis
  Accepted (view paper)
  

  22-Sep-2008 14:20 Migration in Structural Credit Rating Models
  Credit rating migration plays an important role in bank's regulatory and economic capital calculations. Beside the well recognized factors affecting migration, such as macroeconomic conditions and sector, we examine the impact of the underlying credit rating model on the level of migration. We focus on assessing the migration differences of two of the most popular structural credit risk models in corporate lending, the Merton's Distance to Default and the Econometric Model using the Logit function. The results from the simulation study show that migration between the two models can be substantially different for certain parameter values. The consequencies of this claim on a Bank's regulatory capital are significant, as higher mobility under certain macroeconomic conditions will imply higher capital requirements.
• An Overview of Stream of Variation Theory and Applications
  
  Authors: Jianjun Shi The Carolyn J. Stewart Chair Professor
  Affiliation: Georgia Institute of Technology
  Primary area of focus / application:
  

  Submitted at 19-Apr-2008 17:12 by Jianjun Shi
  Accepted (view paper)
  

  22-Sep-2008 14:20 An Overview of Stream of Variation Theory and Applications
  This presentation will provide an overview of research and applications in the theories of Stream of Variation, which is a model-based methodology for monitoring, diagnosis, and control of multistage manufacturing processes (MMP). A MMP involves multiple workstations or operations to produce products. The variations may occur at any operations, which further propagate to the down stream operations. The final product quality is a stack-up, or accumulation of all process variations. Variation modeling, analysis and control for MMP is a very challenging problem due to the complexity in its configuration, variation sources, and interactions among multiple components. In the SoV theory, a state space modeling approach is developed to describe the variation propagation and links the product quality with the process faults in a MMP. This model integrates the product design (CAD) information and the process planning (CAPP) information such as the fixture layout and workpiece setup at each stage. Based on this model, a system level process design, monitoring, sensor-distribution, root cause identification, and active dynamic control are pursued by integrating system and control theory, advanced statistics with engineering knowledge. Besides the SoV research, its applications in industrial processes will also be presented.