INDEPENDENT SCIENTIFIC STUDIES, PROFESSIONAL SURVEYS, OPINION POLLS, STATISTICAL ANALYSIS, GRAPHICS
There is a big need for professional research independent of corporate interests. Consider recently that Gallup Polls shows an Obama approval rating of 43-47%, while a cursory examination of Yahoo comments on articles about Obama seldom reveal more than a 4% approval, even while the articles usually show Obama in a favorable light. Most people believe in the polls, but can you really trust them? Can you trust the mainstream media? Consider safety of vaccines. Nearly all research is carried out by corporate interests or agencies that indirectly receive funding from them. The mainstream has concluded that vaccines are safe and there is no connection to autism, never mind the experiences of thousands of parents who argue otherwise. How about saw palmetto? Does it really work? Nobody knows for sure, but given the state of the art in statistics and experimental design, we should. Many other questions remained unanswered because we cannot trust the results of research provided by the media and funded by corporate conflicts of interest. The goal of Foos Research is to conduct entirely unbiased professional research and provide results to the common man.
CURRENTLY, our first project is an opinion survey to gauge the approval of President Obama and compare that with the Gallup Poll. Those who take the survey will receive a statistical summary of the survey, which also breaks results down according to gender, political and religious affiliation among others. PLEASE take a look at the survey for yourself and submit your answers. Results will be reported via emails and individual web pages in the form of tables and graphs.
Alan Foos has a Master's in Soils from Montana State University. His thesis made extensive use of statistical designs such as complex randomized blocks and regression analysis which gave him a deep understanding of statistical processes. This enabled the development of a theorem describing the connection between treatment significance and consistency of effects across replications. This in turn yields a useful parameter that quantifies the goodness of fit in any experimental design with any number of replicated sets of data. For example, in the above equation, the Foos Coefficient of Covariance (CC) is equal to the Replication Mean Square minus the Error Mean Square divided by the sum of variances for all groups of elements held in common by the replications. The theorem shows how this works for both treatments and replications in a randomized block. For example, the CC for treatments would be the Block Mean Square less the Error Mean Square divided by the sum of variances for each treatment variance throughout the set of replications. In simple terms, a Foos Coefficient of 0 means that there no consistency across replicated elements (blocks or treatments), while a Foos Coefficient of 1 indicates 100% consistency of change for either multiple blocks or replicated treatments in a randomized block experiment or for that matter any group of data with any number of replications.