Pediatric Cancer Rates in Florida (2000-2010)

[raid]

My co-authored manuscript has been accepted for publication, and it can freely be downloaded from here:

http://www.tandfonline.com/doi/full/10.1080/2330443X.2014.928245#.U6WbPvldVKI

The American Statistical Association accepted our manuscript for publication.

 

[raid]

A few weeks ago, it was reported that all playgrounds in Miami had to be closed after finding ashes from a closed garbage burning power plant beneath the sand.

 

[daveleo]

Nice work, Raid !

 

[Spanik]

I haven't read it (yet) and I'm probably not even close to understand in deep or challenge. But statistics has always been fascinating to me. While at the same time something of distrust.

After our exam of statistics we were discussing the asked excersises and found that we (three of us) had gotten to the same anwser up to 3 digits. On the other hand we all three had used a very different method. After that I always have some distrust about statistcs. It can hardly be science if there isn't a "single true way" to get a result. Cannot stop thinking "correlation is no causation" whenever I see something "proved" through statistics. Also I'm always reminded of Feynman's "Cargo Cult Science" where some things are done "because it always hes been done that way" and real factors are not counted. Either because they don't fit the wanted outcome or just because they aren't considered a factor or just plainly unknown to be present.

On the other hand I'm fully aware the a lot of high-energy physics is nothing but statistics. And there is little to doubt that that is real science.

 

[raid]

Statistics cannot prove causation. It can sometimes conclude an association. We depend in many ways on statistics and probability to "show us the way" to the truth. The beauty of statistics is the availability of several models or approaches to study a phenomena. It is not a source of weakness.

Here, we show gographical aras with elevated cancer risks after having removed the effects of different gender and age distributiobs across Florida. The associated probability of having reached this conclusion erroneously is provided. It helps with using common sense planning that is based on statistical evidence.

In Florida (a lawyer obce told me), if a law case is based solely on data, then the probability of an erroneous conclusion may not exceed 1/1000 (ot 1/10,000?) for the case to be
admissible.

The safety of medical drugs is based on clinical trials (statistics). The FDA makes sures that the appropriate statistical mrthods were used.

 

[raid]

The maps show where the cancer rates have been elevated for eleven years. We did not address why.
By the way, close to Miami is the nuclear power plant at Turkey Pt. I do not know if there is any association here or not.