(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
 . . . . . . . . . .



Explaining our death rates while teaching some of you statistical literacy [1]

['This Content Is Not Subject To Review Daily Kos Staff Prior To Publication.']

Date: 2024-06-02  

The way most science is written and the findings presented makes it inaccessible to a majority of our adult population. What I am trying to do here is present some important findings along with what it takes to understand it without assuming you want to know everything it takes to replicate and verify the findings. In other words I am trying to make it understandable rather than show off scientific sophistication.

The above graph uses the most commonly used statistic, r, aka correlation coefficient, to show how the death rates of single year age groups of non-Hispanic whites have changed from 1999 through 2019, the last year before the COVID epidemic. The first dot being down very close to that -1 means the death rate for the people who have not had their first birth day has gone down by very close to the same amount every one of those 20 years. This only tells you the change was consistent and the direction, it does not tell you how much each year of change is. For death rates, that consistent decline is a very good, and for industrialized societies it is now the normal pattern. Next if you look at the dot that is very close to the top and has a hole in it to help you see which one I am writing about here, it shows that that r is close to a positive 1.00, it is actually 0.90, which means the death rate for the 30 year-olds has gone up very close to the same amount every one of these years. The fact that the dots for about fifteen single year age groups close to age 30 means that our early middle aged whites have consistently experienced increasing death rates during the first two decades of this century.

The next graph shows the death rates for our 30 year-olds, the data that produced the correlation displayed in that first dot with a hole in it, for these two decades the death rate for each year is displayed by a dot as high as that year’s death rate on the scale on the left side of the graph.

The general trend of dots showing 30 year-old death rates start with 1999 and 2000 death rates just below 70 and they fairly consistently move up to 128 in 2018 and come down a little for 2019. In the header, the r = 0.90 shows that there is a fairly consistent increase in this death rate by year. The b = 2.43 at the top of the table means that this death rate increased an average of 2.43 deaths per 100,000 each rear from 1999 through 2019. Comparing the numbers for the first and last numbers on the left shows us that this death rate almost doubled during these twenty years. In looking at mortality trends for younger middle aged adults in dozens of countries for this general time period I know that non-Hispanic whites and Native Americans in the United States are the only ones to show more than an occasional small increase in younger middle aged adult death rates in the economically developed part of our world.

But, looking at the relationship between the dots and the straight line suggests this linear, assuming a straight line relationship, is not the best fit. The early and late dots are above the line while the middle dots are all below the line. In effect the upward trend in death rates over time really starts between 2009 and 2010. To get a better measure of what is going on I calculated the r & b for 2009 through. 2019 and got an r of 0.99 and a b of 4.66. This linear relationship covering almost the last half of this time shows an almost perfect relationship, an 0.99 is very close to the maximum possible 1.00 and the b shows almost twice as much of a yearly increase as it did when the data covered the full time period. At the end of this write up I will present a theory about what has caused this shift to a strong increasing death rate.

The regression analysis is designed to look for a relationship that has the same size of a change in the dependent variable, traditionally symbolized by the letter Y, for each change in the measure of the independent variable, symbolized by the letter X. (A remembering hint is that the dependent variable has one leg and must depend on something else to stay standing up while he independent variable has two legs and can stand without depending on something to lean on.)

Following the line of dots showing the changes in deaths rates by single year groups as we got older shows that the dot for age 40 showed the least change for the total period. But looking at the dots shows that the 40 year-olds being in the middle of the downward transition has not been stable, rather it had increasing death rates at first, followed by a decade of declining death rates which was followed by a six year period with substantial increases. This shows the value of looking at a scatter plot of data used to calculate a correlation before accepting a correlation close to zero as showing no relationship, it really only shows that there is almost no linear relationship.

Finally let’s look at the scatterplot for the 70 year-olds, shown with the dot in the hole in it above the word “and” in the top graph.

All the dots for the elderly are so low that their individual scattergrams will very much like this. This is how a plot of these yearly death rates of any group of Americans should look. These two decades have had a very normal changing death rate among most of our elderly.

One of the most unique things about health care in the United States is that people of different age groups have health care paid for very differently. Health care for the young is primarily covered by private health insurance but about a decade ago the domestic violence coalition lobbied for health coverage of the young who were uninsured. All the states have put in a way to provide care for the uninsured young for the primary reason to keep abusive husbands from using the threat of canceling health insurance for their young if the abused wife/mother leaves the abuser. It turns out that the system put in place has so substantially reduced the death rate among the young has had to save many more lives than its intended purpose. The over all effect is that almost all of our young have access to health care.

The middle younger middle-aged are unique in the modern would among economically developed countries. This pattern of increasing death rates exclusively in the middle age is just about as deviant and bad as anything can be. The major cause of this is not directly measured but is occurring only among those private health insurance dominates and it has accelerated after the private health insurance system put in place required prior approval of health care before it would pay for it. And this pattern of increasing death rates ends between ages 55 and 65 when the very ill can apply for early Medicare in most of our states. I have no idea why it goes up so much between ages 48 and 54,

[END]
---
[1] Url: https://www.dailykos.com/stories/2024/6/2/2240827/-Explaining-our-death-rates-while-teaching-some-of-you-statistical-literacy?pm_campaign=front_page&pm_source=more_community&pm_medium=web

Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/