Asee peer logo
Displaying all 2 results
Conference Session
Issues and Opportunities in IE Education
Collection
2009 Annual Conference & Exposition
Authors
J. Eric Bickel, University of Texas, Austin
Tagged Divisions
Industrial Engineering
, T(r | p)  T(p | p) for all r  p andT(r* | p)  T(p | p) when r* = p. [1-4] Many strictly proper scoring rules have been developed.Three of the most popular are given below.Quadratic (Q): Qi (r )  2 ri  r  r  [1,1] (2)Spherical (S): Si (r)  ri / (r  r)1/2  [0,1] (3)Logarithmic (L): L i (r )  ln( ri )  ( ,0] (4)The range of possible scores differs considerably. For example, logarithmic scoring holds thepossibility of an infinitely negative score. While this may seem like a defect, we will argue thatthis feature is a benefit of log scoring. Any linear
Conference Session
Issues and Opportunities in IE Education
Collection
2009 Annual Conference & Exposition
Authors
Marlin Thomas, Air Force Institute of Technology
Tagged Divisions
Industrial Engineering
starts with the multidimensional definition of quality, Q ? f (Q1 ,..., Q n ) with the nelements that correspond to those factors and features that relate to how the products aredesigned, developed, produced, and used by customers. Functions involving design, production,and service have differing effects among the n dimensions, producing large impact on some andvery little on others. This of course will depend on the particular type of product. The nature ofthe dimensions makes it difficult to establish an overall simple measure of the state of quality,with some dimensions being quantitative while others are very subjective. Warranty feedback, W ? h ] g1 (Q1 ,..., Q n ),..., g m (Q1 ,..., Q n )_ does however, provide an overall weighting of