Brad's Hypothesis Generator
I created the hypothesis generator in 2005 to increase my knowledge about the health subjects I was interested in at the time. Since it's written as a javascript app, it is easily shared with others who value such information. There's a link at the end.
The lists of numbers are probability values of the correlations between data sets, but they are presented as the reciprocal of traditional p values. Instead of expressing a 1 in 20 chance as p = 0.05, it is expressed as 20 which is much easier for a non-scientist to understand. When a mouse is hovered over a number, or the number is tapped on a phone, a popup provides details about the correlation.
When the values are positive, the datasets are in alignment. For example, "higher" smoking rates correlate with "higher" death rates. When the values are negative, the datasets are inversely correlated. For example, "more" green salad correlates with "less" obesity.
The list is 130 datasets long...
The hypothesis generator will look like a treasure hunt to some people, but like an absolute bore to most. If you're one of the rare people who find it intriguing, you can keep a copy by placing the two pages hypothesis-generator.html and engine.html into a local directory. The second page, not shown here, documents the data sets and can calculate each probability value beyond the million iterations of the main page.
This resource is a model for a much more useful and powerful version that I'd like build, as described in The Mechanism of Aging. That would require a million dollar survey to create the datasets, so unless a miracle happens this version is as far as the project goes. In any case, the current resource is useful for illustrating some thought-provoking correlations.
My advice on using this resource is to recognize that cause and effect can be convoluted. For example, sunburns are strongly associated with the states that have the least sunny days. That seems wrong, until we consider that people lose more tan during the long winter in colder climates, and are therefor easier to burn during the short summer. There are many puzzles like this that are less obvious, so don't imagine theories when all that the data allows are hypotheses. Nevertheless, there's a lot of knowledge to be gained from pondering the evidence.
The hypothesis generator is not designed for a small screen.
https://mystic-sagery.com/hypothesis-generator.html