10/15/2005
Data & Hypotheses

When you open yourself to trust foreign sources that are broadly and globally minded, your data and conclusions may be swayed when the bias is not obvious. The problem as I see it has at least two relevant faces. One is the political bending of data released, including their suppression in whole and in part. The second is the relevancy of variables that actually determine the true value of statistics and information used in predictive models.

I find that if you watch national US television news or read the New York Times or Washington Post you know far less about themes like European, African and Asian issues in time or early enough, as compared with the BBC  or French media -- which I do follow almost daily. I find the bias of BBC TV and radio very consistent and virtually predictable (especially -- but not only -- when you cross check) against:

1. anything U.S. or Republican;
2. Israeli initiatives or wishes in the Mideast conflict;
3. top dogs in economics and industry, including a consistent attempt to paint blacker than the devil their interests and leadership (particularly when not British);
4. new types of commerce, financial expansion and local entrepreneurship;
5. demands and suggestions to create productivity and self policing in less developed countries.

The BBC source is refreshingly  very consistent in preferring the stated ideological predisposition in favor of simplicity in reporting bare facts. It is committed to black-and-white, missing shades of gray and colors altogether. And it misses, seemingly accidentally but actually permanently, evidence or need for applying the rules of evidence.

However, I do need to check the BBC, Arabic language media, many terrorist websites around the world and numerous other propaganda sources for three reasons

1. to develop hypotheses
2. to make sure I do not miss an important scene or perspective
3. to consider their alternative explanations and predictions before I come with my own opinion and checks.

It matters not at all where you live in this age of Internet and search engines; you can access sensitive and complete information if you really want to know. If it does matter where one lives, one should prefer to be, no insult or diminution meant, a barber or a fashion director, as the specialty rights and wrongs are, while not less biased on getting valid and reliable causality, faster in reaching targeted practicing professionals.

Having lived in Israel, Japan, Korea, Brazil and Papua New  Guinea I have always found information filetered through cultural prisms to be less accurate but more enriching for analysis. Getting to ask the right questions is where knowledge starts. Getting to have an open mind and an independent pool of greatly informing models and hypotheses is an exercise our chair and many of our specs consider imperative and for good reasons.

There are more serious issues for the student of data patterns: too often we assume that we know the research universe relevant for prediction. For example, Chinese-made cars are sold in Europe -- mainly from a Belgian base of sales -- at about half the price while duplicating original European and Japanese models that are either very prestigious or economical. The numbers are growing at such a rate that this phenomenon will become noticed in less than two years. The overall consumption of energy in China and India is not reported fully, as new manufacturing plants, including those in certain remote areas, do not report their total energy needs in 2006 or 2007. The growth in production of consumer goods in these and other populated countries seem to predict new types of burden on the national and global energy accounts. Very often these data are collected or/and known to the energy giants and analyzed by them before public or external government  is aware of coming changes.

With  these factors in the background, I also use a few search engines that are mathematically  and statistically able to help look for patterns where you find none or only glimpses and, hence, create distal knowledge. Often I find that numbers and quantitative measures taken for granted, are off or wrong at the source or sourcing  by looking for covariance or missing data that should be easy to locate, given what we know about other such studies of the virtually same universes elsewhere. Often enough these are not real issues -- but when this extra care is applied the following, in  particular, require a second look:  energy, food consumptions, agricultural products (especially genetically based), certain financial reporting trends and utilization of related statistics, capital reserve deposits and metal extraction. Then, it appears that very sophisticated analysis using the less-than-optimal sources, particularly  when not comparing a few independent sourcing, results in bigger sins, misleading a researcher or his research consumers. Again it can and does happens almost everywhere, not only in America.

Prof. Joseph D. Ben-Dak is an expert in international security, responses to terror, technology and global politics. He has served in numerous high-level international posts at the United Nations and other organizations. He holds a doctorate in Organizational Sociology and Research Methods from the University of Michigan, Ann Arbor.