We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

Blog: Dan Power Subscribe to this blog's RSS feed!

Dan Power

Greetings to all of my friends who work in the area of computerized decision support. This blog is a way for me to share stories from my encounters related to decision support, to comment on industry events, and to comment on other blogger's comments, especially those of my friends on the Business Intelligence Network. I'll try to state my opinions clearly and provide an old professor's perspective on how computers and information technology are changing the world. Decision making has always been my focus, and it will be in this blog as well. Your comments, feedback and questions are welcomed.

About the author >

Daniel J. "Dan" Power is a Professor of Information Systems and Management at the College of Business Administration at the University of Northern Iowa and the editor of DSSResources.com, the Web-based knowledge repository about computerized systems that support decision making; the editor of PlanningSkills.com; and the editor of DSS News, a bi-weekly e-newsletter. Dr. Power's research interests include the design and development of decision support systems and how these systems impact individual and organizational decision behavior.

Editor's Note: More articles and resources are available in Dan's BeyeNETWORK Expert Channel. Be sure to visit today!

Some readers may have noticed that my turnout forecasts for the Iowa Political caucuses were extremely inaccurate. "In 2004, 125,000 Iowa Democrats caucused. Turnout for 2008 should be similar, about 230,000 total." Well predicting demand using just a few data points generated 4 years apart is definitely a mistake. In reality, about 239,000 Democrats and 120,000 Republicans caucused. Why the enormous error and what is the impact of "bad" forecasts?

Why? Many new attendees encouraged by the well organized campaigns, extensive media barrage, no incumbent candidate, good weather .... Turnout was also very high in New Hampshire.

What is the impact of a "bad" forecast? Resources are misallocated, polling can be misleading, and perhaps worst of all, behavior may be impacted, voter, campaign staffers, media.

Could we have avoided the horrible forecasts for turnout? Perhaps, combining historical data with voter intention data from phone surveys could have helped.

Could we have made more accurate projections of votes for winners and losers? Maybe, but the electorate is undecided and some of us decide at the last moment. Also, in voter preference polls, people don't always tell the truth. The margin of error is very high right now for many reasons (certainly more than statistical sampling error).

The best demand forecasting tools require that the future is an extrapolation of past behavior and that intentions capture future behavior. ... A very weak set of assumptions this year.

Regarding Second Life:

Group Notice From: Leinad Meriman aka Dan Power

You are cordially invited to Decision Support Systems Workshop #3 with host
Richard Hackathorn (Hack Richard) for the topic "Serious Games in Virtual
Worlds: The Future of Enterprise Business Intelligence". It will be held
Friday, Jan. 18, 2008 at 8am PST (16:00 GMT, 17:00 CET) at TechTalk@SL
Discussion Center. Everyone is invited - the workshop is free and open to the broad SL audience.

check http://www.b-eye-network.com/view/4163

Also, please visit the new Decision Support World global headquarters in Second Life. It is still under construction, but you can get a sense of what is planned.

Posted January 14, 2008 7:25 PM
Permalink | No Comments |