Statistics

[Ronald Zeeman]

As many you may have noticed I like playing with stats.
Years ago I put together a blank statistics spread sheet.
If any one here is interested it is a simple individuals and moving range.
I am willing to send copies, Excel format. If you want a copy you will need to send me an E-mail address privately, on the message part of the this forum to send it too. Just plug in the data of interest and it does the calculations automatically, it also contains a Z-table calculator.

 

[Neonatal RRT]

...and I thought I was the nerd.

 

[Gerald Wilgus]

I am working on a chi-squared run chart for frequencies in aquatic invertebrate sensitivities based on Hilsenhoff tolerance values ordered by quartiles. The objective is to detect any shifts that indicate a degraded environment as soon as possible.

My real passion is Statistical Design of Experiments. I had devoured the book by Box, Hunter & Hunter. And I still have a Western Electric Handbook of Statistical Process Control.

 

[Ronald Zeeman]

I took a course in design of experiments, nice method a little to sophisticated for most stuff that I do.

 

[Ronald Zeeman]

Here is an example of what you can do with data, any dataset, I did a few this years ago, mathematically turned curved lines into a straight line then to control chart. Sports example, additionally I joined weight watchers years ago. Tried to convince them this would be useful for their members, fell on deaf ears.

 

[Gerald Wilgus]

Statistical graphics, done properly can reveal a lot. I once was tasked with validating temperature controlled rooms for pharmaceutical stability studies for control and uniformity. Data recording points up the wazoo. The first thing I did was plot box and whisker plots of temps over 24 hours at each location. First thing I noticed was a set of points across a section of a room that was significantly lower than the rest of the room (no overlap). Investigating, I found that because of space restraints an insulated coolant line had been run across the air supply plenum above that region. Such analysis and an analysis of control as external conditions changed and with load hysteresis, let me design a monitoring plan that saved the company $400,000 a year. Did I see any of that? Nope.

 

[Ronald Zeeman]

I was know for ignoring management team, finding clever ways around them and getting My Ideas accepted. When they worked watching how fast they take credit for it. incremental changes work much better than one big change every one need to buy into. I visualize the big change in my head, plan how to get there slowly, incrementally follow my plan say nothing to any body. just quietly implement it. Then one day it works. bigger company buys mine. no one the wiser. Sort of like the frog in the hot water.

 

[Gerald Wilgus]

I was know for ignoring management team, finding clever ways around them and getting My Ideas accepted. When they worked watching how fast they take credit for it. incremental changes work much better than one big change every one need to buy into. I visualize the big change in my head, plan how to get there slowly, incrementally follow my plan say nothing to any body. just quietly implement it. Then one day it works. bigger company buys mine. no one the wiser. Sort of like the frog in the hot water.

In my last position I was fortunate. My work in support of quality, especially with sterilization processes and pre-inspection audits gained me a lot of credibility and influence. I was the fly in the ointment, though, among product managers who wanted to import from China. I had Engineering (who I could trust) test the output from those sources and give me the data. Many times I could explain just how the processes were out of control or post process sorting took place to yield material which took up the entire tolerance interval. They disliked to see me coming, but I felt that I worked for the patients using our products, not those managers.

 

[Ronald Zeeman]

I just found a law in statistics I never was aware of it's called the 0 1 law. explains something in Themo dynamics. I was unaware of until I watched my great course lectures. on phase diagrams interesting.

 

[Ronald Zeeman]

I put this together years ago on aging the more I use it the more accurate I notice it is. We are bringing in more and more immigrants to my country and this explains why. Simple z-table calculations easy to use. In a few more years a number of countries will be competing for immigrants, not necessarily educated as us baby boomers pass away.

 

[Ronald Zeeman]

Found out yesterday one of my nephew hobbies is statistics he likes to do analysis get the results published. Told him I would send my covid data in he was really excited. His mother is a trained actuary and wife works for police force crunching statistical crime data, I hate sitting on my covid data, should be public.

 

[Ronald Zeeman]

My nephew is now checking out the stats I sent him He really know his stuff, I collected the data finally have someone who will do a proper analysis of it may be write and publish some papers on it. I have two nephews that are amazing with math.

 

[Ronald Zeeman]

Interesting conversation with nephew he sate correctly each individual case is not statistically independent of each other, I stated I view each sampled area as a liquid, where the number of cases per unit time is equivalent to concentration of the liquid, changing over time, Visual thinking over rigorous math. Then I seek inflection points.

 

[Ronald Zeeman]

Nephew was surprise bell curve formed I was not again liquid analogy two liquids of different concentrations mix eddie currents form virus spreads unevenly clumping looks like eddie currents again visual thinking. Looking back now makes perfect sense. I can just see this. Natural process of course it would bell curve. why would it not. Hopefully my nephew gets me after all his dad is a fellow Aspie,

 

[Ronald Zeeman]

The effects I'm seeing are not dependent of statistical independence rather on emergent effects due to infected people interacting with each other like eddy currents in a two liquids mixing. I fortunately can visualize this without rigorous statistical analysis.

 

[Ronald Zeeman]

I put stats together years ago on rate of living. starting to scare me how accurate they friend of my wife and I just turned 65 yesterday, now at hospital with staphylococcus infection, at my 65 ten days later I had stroke another friend of my wife was in hospital at 59 another inflection point, three older ladies we knew passe at 93 an inflection point. Only probability, bothersome. My wifes best male friend hospitalized at 70 looking for long term care, inflection point my current age.

 

[Ronald Zeeman]

Just noticed DR. Yang came up with YAng-mills theory passed on four sigma 103 years. Was oldest living physicist.

 

[Ronald Zeeman]

My nephew sent back this:

It's an autocorrelated process, so applying assumptions of normality
and independence really doesn't make sense.

I've created a really simple time series model in the attached file,
where adjusting the replication rate (i.e. lockdown measures etc.) can
recreate the same waves that we see in the data. If you had a random
normal process doing that it would just look like noise instead.

I responded with this .

Interesting you see it as stats issue I'm a highly visual thinker , being on the autism spectrum. So think of this as an engineering issue two liquids mixing the daily sampling is measuring the liquid density per moment time one liquid has no virus the other lots the density changes as eddy currents form each person is like a drop of each liquid so human nature determines when and how they will mix and human genetics will determine the individual density of each drop. So the real question why did it form a bell curve . I can visualize why it would as a I AND R chart works this way.

Fortunately my nephew is an engineer. Top student in his class . Maybe this may give insights into Navier Stokes .