Re: Correlation is not causation, yet a fascinating observation jpeg

3
All the things we've posted about. We've said that it's unlikely a person with a stable job and promise of a decent retirement, access to health care and to education, with access to affordable housing and a variety of foods, to recreation and life in a stable and orderly community--these people rarely run out and shoot up others. Beer is just icing on the banana.

CDFingers
Crazy cat peekin' through a lace bandana
like a one-eyed Cheshire, like a diamond-eyed Jack

Re: Correlation is not causation, yet a fascinating observation jpeg

4
Interesting, yes. And likewise there are plenty of other metrics to plot at the same time. However, how are we measuring strictness of gun laws? That seems a bit arbitrary at best, but I bit and tried to figure out what was meant…

But as it is, I’m unable to find the source. It says, “crimadviser.com” which does not work. Likewise the correctly spelled, “crimeadviser.com” is just a GoDaddy parked domain name. Tossing it in to Google only brings up similar and other graphics and crime, but none seem particularly well thought out or linked to actual substantive data, analysis, or at least a working source.

So while my feelings about common sense laws (which is a whole different discussion) might align with this (depending upon how it’s defined), I can’t put this at factual information. If someone has better information, please offer it.

The cdc.gov firearm mortality rate by state is interesting, though there are still plenty of ways to correlate it as stated (education, income, etc).

https://www.cdc.gov/nchs/pressroom/sosm ... irearm.htm

That’s probably already discussed here I would guess.

Anyhow, cheers all. And happy thanksgivings.

Quo

Re: Correlation is not causation, yet a fascinating observation jpeg

6
I don't take these charts seriously, no footnotes or exact references to data. Anyone can build a chart to justify their position. Crimadvisor.com was a Brady Campaign site that doesn't appear to exist any longer, Safehome looks like a DV group not sure. We know that 2/3rds of gun deaths are suicides and the chart doesn't break it out. What do they consider "strict gun laws", the murder rate in CA and in TX is about the same. Another simplistic chart that wows people on the internet.

IIRC there was another thread recently on the same topic.
"Everyone is entitled to their own opinion, but not their own facts." - Daniel Patrick Moynihan

Re: Correlation is not causation, yet a fascinating observation jpeg

7
I'm a political person. One of the reasons I'm here is to enrich the political situation for liberal gun owners. Liberals enjoy the reputation of being anti gun, so I find it helpful to post these things so the Club can support policies that enrich us in that way. I may have to cough up cash for my own software.

Now, when one goes to the links I posted and toggle between the OP jpeg and those numbers, there is easy correlation to be found. The hard thing is to determine the "strictness" factor.

Just this morning we see another domestic violence incident in TX, resulting in two deaths. TX is 95% privately owned, and its lege is gerrymandered such that only red folks get in. CA is a tad different, with just under half being owned by the Feds, with the rest about split between state and private. Moreover, we have a very fair redistricting process, contrary to TX. CA has ten million more people with a gun death rate of 8.5 per 100,000, yet TX has a rate of 13.7 per 100,000. TX has a higher poverty rate. The biased reader might look only at guns. But when we factor in the socio economic factors, we see that TX is cool if you're a hereditary land owner, but if you just drive their crappy roads, have to send your kids to their crappy schools , struggle with crappy health care, or try to get anything done without owning a local politician, you're SOL. When we look at the whole picture, the objective reader might have an "ah-ha" moment. Montana's gun ownership rate is the highest, but their gun death rate is, well, no where near either TX or CA.

This is why I focus on root-cause mitigation so frequently. It's not the guns. It's the circumstances of the state. Here are some links I used in the seven minutes required to make this post. It's quite easy to dismiss a groovy chart. But upon careful observation, there is correlation. Not causation. Recall the title of this thread.

https://www.census.gov/quickfacts/fact/ ... /PST045221
https://gis.data.ca.gov/search?categori ... 0ownership
https://worldpopulationreview.com/state ... p-by-state

We read that Biden wants to ban "assault weapons." All gun owners are against it. I want to show why we're against it.

CDFingers
Crazy cat peekin' through a lace bandana
like a one-eyed Cheshire, like a diamond-eyed Jack

Re: Correlation is not causation, yet a fascinating observation jpeg

8
We're both on similar tracks. A simplistic chart like that one only looks at one possible factor in gun deaths, it doesn't look at the many variables beside gun control laws that also contribute to gun deaths. Where do most gun deaths occur, homicide or suicide, gender and age, socio-economic level, personal, gang, drug or other connections... And to try and isolate all causative factors like gun laws, they need to be analyzed within a state and compared to other states.

I could make up a chart showing that crime increases with population density, like adding more and more rats to an enclosed area will create conflict. But there are other factors such as poverty, unemployment, lack of housing...but cities also have wealthy areas with little crime. Factors need to be analyzed to determine how important they are within the totality.

People especially politicians like to grab the most simplistic answer that they can use in campaign stump speeches. They are driven by the media and polling, Biden is out talking about another AW ban because the media is beating the drums on the recent shooting.

I don't agree that CA has the fairest process for redistricting. It's mostly handled by the CA Auditor's office, the CA State Auditor is a political appointee. S/he selects the auditors that review the applications for commission appointments and conduct interviews and create the lists. Sorry but I'd rather have a partisan legislature make the decisions, at least the partisanship is in the open.
https://www.wedrawthelinesca.org/about_us
"Everyone is entitled to their own opinion, but not their own facts." - Daniel Patrick Moynihan

Re: Correlation is not causation, yet a fascinating observation jpeg

10
Without a data table, I had to eyeball the position of each data point before running the correlation coefficient. Also, this data can't be recent. So, grain of salt.

With strictness index as the x axis, the r^2 value is 0.58.

That correlation suggests that 58% of the variance in per-capita gun violence is related to strictness index.

However, it does not tell us whether gun violence is an inverse function of strictness index, whatever the fuck that is, or whether the strictness index indicates gun regulation is increasingly strict following increases in per-capita gun violence. Reasonable people can differ on that interpretation given the limited data, but the gun control skeptic would argue that legislation is a lagging indicator - that strict gun regulation is unlikely to precede and prevent problems with gun violence, but is more likely a response to an increase in it.

This problem is much more amenable to a computer-assisted multivariate statistical treatment of the sort our Dr. Yamane might be interested in. It has been a long time since I was forced to take a class in black box math. I suspect per capita GDP, percent incarceration, Gini coefficients and other socioeconomic parameters would score higher correlations.

Re: Correlation is not causation, yet a fascinating observation jpeg

11
Wings
...the gun control skeptic would argue that legislation is a lagging indicator - that strict gun regulation is unlikely to precede and prevent problems with gun violence, but is more likely a response to an increase in it.
I agree


Gun death rates are different than gun homicide rates - gun death rates include suicides. If you look at Montana south to New Mexico that's referred to as the Suicide Belt, it actually starts with Alaska. Highest rate of suicides are among Native Americans and whites.


Image
"Everyone is entitled to their own opinion, but not their own facts." - Daniel Patrick Moynihan

Re: Correlation is not causation, yet a fascinating observation jpeg

12
Yeah. As gun owners, suicide numbers mean a lot more to us than they might to non-owners. Obviously some of the factors driving gun suicide are radically different from homicide, but they are typically deaths of despair - motivated at least in part by inadequate access to health care, mental or otherwise, economic injustice inequality, and a lack of hope for the future.

I'm reading about the Walmart night manager shooting up his store in Chesapeake and how he legally purchased the handgun the day of the shooting. I think waiting periods are an annoyance, yet I've never bought a gun on impulse. If my annoyance at adding 5-10 days to a purchase that I probably spent months considering - speaking for me - saves a life from someone on the brink of an impulse-driven murder spree on a bad day? I'm in. Start talking magazine restrictions and my eyes get a little squinty. Bans? Pft. We allow full-auto weapons with tax stamps. Couple weeks back I went to the range and immediately recognized a fellow shooter was there by the sound of short bursts of automatic fire from the pistol range. The steel was safe, just saying. I don't want randos banging full auto during drive-bys - but bans don't seem to be the answer. I think the NFA has actually struck a balance between legal access and federal regulation that seems to work, philosophy aside.

I sure AF wish we had more data and am deeply annoyed that the NRA has stood between us and any sort of dataset that would inform this discussion.

Re: Correlation is not causation, yet a fascinating observation jpeg

13
It's not that the OP's chart is wrong--the correlation is real, but there are, as pointed out, many possible variable that lead to that correlation.

First, I'd look at Zach Mortensen's thesis that income inequality is THE most significant factor in gun deaths (and I'd hypothesize that it's true of suicides, too).
Liberal states tend to be more generous with their social welfare programs, all of which numerically increase the income of those who receive them. Schools, social welfare programs, health care access, access to contraception, laws to combat red-lining ALL exist in liberal states. Our organization's Root Cause Mitigation theme is more likely to be presented in various means. I have cousin who works in NYC as a lawyer in Restorative Justice, and liberal states are much more likely to fund RCM's similar program, Hospital Violence Intervention Programs, or HVIP, which is PROVEN to lower gun violence....until legislatures stop funding it, as they are more likely to do in Red states.

I'd look at the effectiveness of gun access control--the Walmart shooter bought his handgun that very morning. Since his record was clean he might WELL have been able to buy it in my state of NJ, but he couldn't have been impulsive about--and impulse control may keep some shooters from acting out. I think that the better and more universal background checks are may WELL prevent some shootings. The Texas Church Shooter a few years ago obtained his AR legally, because the USAF never sent his history to N.I.C.S.

States that don't give a flying fuck about people, where pols just want votes so they say and do what the yahoos want, results in a series of laws that fuck up their state in all kinds of ways. Remember when Kansas lost over 700 teachers en masse and they were writing books "What's the Matter With Kansas?" Well, I posit that the same things that were wrong with Kansas that lead to a whole SLEW of troubles also lead to increased gun violence.
"Even if the bee could explain to the fly why pollen is better than shit, the fly could never understand."

Re: Correlation is not causation, yet a fascinating observation jpeg

15
CowboyT wrote: Fri Nov 25, 2022 9:15 pm I look at charts like the OP's chart the same way I look at this chart, and for similar reasons.
Image
Praise Be His Noodly Name. He boiled for your sins, don't forget.
Image
What your missing is that the OP's chart is attached to something UNDERNEATH the gun control laws, namely that the states that implement do many other things likely to lower gun violence, things that states with higher rates of gun violence DO NOT DO! Simply arguing that correlation is not causation is simply that: Simple, in fact, overly simple, like a crowd chanting something like "Lock her up!" or "Stop the Steal" to drown out meaningful debate.

Numbers of pirates are not surrogates for an underlying cause of changes in pollution, but gun control laws ARE a surrogacy for why those states DO have lower gun violence rates, because that same liberal impulse that seeks to lower gun violence by gun control also takes other steps that can and should lead to lower gun violence rates.

There ARE appropriate times to use the correlation analogy, especially when there are insufficient data points being made. When I worked at Anti-Trust over 30 years ago, lawyers and statisticians tried to argue a connection using only 7 data points. The DOJ statistician showed that the price of gilts and sows (pigs) were just as good a predictor of the dependent variable, blowing their case apart.
"Even if the bee could explain to the fly why pollen is better than shit, the fly could never understand."

Re: Correlation is not causation, yet a fascinating observation jpeg

17
I think it a worthy challenge to solve the conundrum of mass shootings in America, a country with the 2A. Easily we may correlate all kinds of facts, like geographical location, age, income, race, and so on. But we're dealing with humans, a fuzzy science at best, and the best-selling Pet Rock at its worst. We really can't get a precise answer like 1.6180 or something when dealing with humans.

We can, however, detect that almost all mass shooters are men, and most of them fall into the age range of between 17-30 or so. Between 17 and 30 I always had a woman in my life. Is that why I never ran out and shot up the local Dairy Queen? I'll never know. As it so happens, the only state where prostitution is legal is Nevada, with two mass shootings in the last forty years.

https://www.statista.com/statistics/811 ... -by-state/

I'm not saying that getting laid regularly between 17 and 30 saved my local Dairy Queen--it got me a son, though. Yet prostitution might be a thing to look into, to regulate and tax. What could go wrong? What could go right is that women (likely the gender of more than half of prostitutes) would enjoy more autonomy over their bodies, and incels could spend money to stimulate the economy in ways different to how they do now. I've known one unregulated gay male prostitute friend in my life, not in the Biblical sense, who never shot up anything. He was a Vietnam vet who died of AIDS in the mid 80's. May peace be upon Ken.



Food for thought.

CDFingers
Crazy cat peekin' through a lace bandana
like a one-eyed Cheshire, like a diamond-eyed Jack

Re: Correlation is not causation, yet a fascinating observation jpeg

18
CDFingers, that's what social sciences all do: Model human behavior. The modeling is usually quantitative. I worked as graduate assistant for a guy modeling the economics of criminal behavior. He had told a class that all human behavior could be modeled in Econ. One kid asked "Can you model crime?" and that sent him off in a new career direction. We modeled effectiveness of treatments on recidivism, contrasting economic vs violent crimes, changed measures of recidivism etc.
Criminologists were modeling behavior from different perspectives.

Sociologists model, as do Political Scientists, Psychologists, and Economic Historians (Karl Marx's other great achievement was viewing History from an a vast economic viewpoint).
"Even if the bee could explain to the fly why pollen is better than shit, the fly could never understand."

Who is online

Users browsing this forum: sikacz and 3 guests