 |
Our Clients
Request Info
|
|
About the Company
|
|
Statistical Innovations, a Boston-based firm which specializes in innovative applications of statistical modeling, was started in 1981 by Dr. Jay Magidson. Dr. Magidson identified a real need in the marketing research, biomedical research, and other arenas to improve upon the way that data analysis was being conducted. The assumptions underlying the methods being used at that time did not hold and therefore were providing misleading results.
In the mid 1970's, while completing his Doctorate in Managerial Economics and Decision Sciences from Northwestern University's Graduate School of Management, Dr. Magidson began studying the work of Professor Leo Goodman, one of the pioneers of our time.
Professor Goodman's important contributions to statistical modeling include the development of log-linear models, association models, and the methodology for estimating latent class models. Dr. Magidson felt he could bring the ground breaking techniques, models and methods of Professor Goodman into real-world applications which would result in tools that would help companies solve their modeling problems more effectively.
From these beginnings Dr. Magidson has continued to provide businesses with innovative "outside the bracket" thinking, encouraging them to use creative inference to solve problems rather than the application of rigid statistical formula.
Statistical Innovations Inc. was first to:
- introduce a full Windows-based program for latent class cluster, factor, and regression models
- provide user-friendly datamining software to segment and explore one's database, SI-CHAID® and SPSS CHAID (#1 datamining analysis software)
- show how a latent variable model could be used to evaluate the effects of an intervention, now a common practice (Magidson, 1977)
- present a published collection of the classic articles on categorical data analysis by Professor Leo Goodman to the public (Goodman with Magidson, editor, 1978).
- present a published collection of the classic articles on LISREL-type models by Professor Karl Joreskog to the public
- recommend the combined use of CHAID and logit modeling in Database Marketing (November 1981 Journal of Marketing Research), a practice that has been since adopted by advanced model developers
- show that the Monotonic Regression Model could be used to improve upon discriminant analysis for the classification of Fisher's classic iris data and other ordered groups (See GOLDMineR® , 1998, Chapter 11).
To learn more, contact
Will Barker
Sales & Marketing Manager
Statistical Innovations
375 Concord Avenue Belmont, MA 02478-3084
Phone: +1.617.489.4490
Fax: +1.617.489.4499
Email: will@statisticalinnovations.com
|
|
New release: Latent GOLD® 4.0 >> 
