Newt Gingrich exemplifies just how unscientific America is (Synopsis)

“When an issue comes up, and it can be decided based on FACTS — on DATA, on evidence, on things you can MEASURE and QUANTIFY — is it ever okay to decide that issue based on your feelings rather than what the facts indicate?
It isn’t just Newt Gingrich’s comments about violent crime, meant to incite fear even as CRIME RATES CONTINUE TO DROP nationwide, but a symptom of a larger epidemic in the world.”

Haven’t read your Forbes piece yet. Perhaps it has facts and data to support your statements.

However, here’s some of what I’ve read:

“Federal Bureau of Investigation Director James Comey has again drawn the wrath of the White House for calling attention to the rising violence in urban areas. Homicides increased 9% in the largest 63 cities in the first quarter of 2016; nonfatal shootings were up 21%, according to a Major Cities Chiefs Association survey. Those increases come on top of last year’s 17% rise in homicides in the 56 biggest U.S. cities, with 10 heavily black cities showing murder spikes above 60%.”
http://www.wsj.com/articles/the-nationwide-crime-wave-is-building-14640…

So sn, you have no understanding of statistics either? Why am I completely unsurprised?

Haven’t read your Forbes piece yet.

An all-too-common introduction to an S.N. comment. At least I didn't have to type "n*ggers, amirite?,"* which I could see coming before reaching the final clause.

And no, I have zero interest in feeding this moron's desire to threadjack with brain-dead** political posturing.

* "Neil deGrasse Tyson is the Al Sharpton of science," etc.
** I think I may have a screen shot lying around of its Disqustink comment consisting solely of "That B I T C H should be in jail!" (modulo exclamation points).

I's George Luca's fault, "go with your feelings..."

Now that may seem like a huge overreach, but in fact our entertainment culture really is pushing us in this direction. We don't want to look stuff up, we want the instant gratification of just knowing (feeling) what is right. Movies help a lot, we know someone is a villain usually before he does anything nasty, its something to do with how he looks and how he moves, that gets our gut telling us he's creepy.

So we are training our mental reflexes for hours a day, to make snap emotion based judgments, and the habit carries over.

For the world that Gingrich occupies, a world of promotion and marketing, his observations are correct. Your product may be a piece of junk, but if you can create an exciting image, it will sell like hotcakes. Probably more people make their living out of marketing style endeavors, than make their living trying to wrest truth from nature, so for them his mode of thinking is correct (promotes personal success).

SN: crime rates are dropping nationwide. At the same time yes there can be local increases in specific areas. Gingrich discussed that exact topic with Camerota. However, he then tries to say that looking at national statistics is a liberal thing while the GOP is representing peoples feelings (direct quote: "The current view is that liberals have a whole set of statistics which theoretically may be right, but it's not where human beings are. People are frightened.".). And in the very next exchange, he reiterates that he doesn't really care about the facts, he cares about how people feel about crime:

CAMEROTA: "...This is the FBI statistics. They're not a liberal organization.

GINGRICH: No, but what I said is equally true. People feel it.

CAMEROTA: They feel it, yes, but the facts don't support it.

GINGRICH: As a political candidate, I'll go with how people feel..."

So, yes, (a) crime rates are dropping nationwide, (b) Newt implies that he actually accepts this to be true, but then (c) basically admits that he, as a conservative politician, cares more about catering to people's feelings than what's factually true about the world.

The really sad thing is, there is plenty of room to have a conservative/liberal policy debate about things like incarceration as crime rates go down while at the same time some evidence suggests its concentrating in larger cities. There's a policy debate to be had about what international treaties we should make to control or reduce anthropogenic climate change. Admitting that crime is going down and AGW happens does not necessarily imply that a liberal policy response is the correct policy. But we will never have those substantive, important policy debates as long as the main conservative party in the US decides to use fearmongering and 19th century know-nothing political strategies to drum up votes.

SN: crime rates are dropping nationwide. At the same time yes there can be local increases in specific areas.

You just gave him the reason behind my comment @#2. I was waiting for his vapid attempt to explain how his numbers invalidate the national trend. I was betting on the reason having something to do with all the people who willfully submitted to execution rather than deny that they had seen the resurrection.

I think that poll is an example of why polls are a terrible way to understand the world. It's a stupid question to ask - how is some random person able to know if the crime rate is down, up, or anything.

"I think that poll is an example of why polls are a terrible way to understand the world. "

Polls like this one are done to gauge perception , nothing more.

This article is deeply disappointing. Everyone is dumber for having read it.

Here is a hint: We didn't stop collecting crime data in 2013.

The truth is after a long and steady decline in violent crime rates since the 90's, the crime rates bounced and are on the rise. In the first half of 2015 violent crime was up 1.7% nationwide over the same period in 2014 [1]. When you look specifically at homicides in the nation’s 56 largest cities the rates were up by 17% in just one year!!! [1]

New crime statistics show the 2015 spike wasn't an aberration. The trend is accelerating with 2016 numbers worse in every category. [2]

Homicide is up another 9.1%
Rape is up 4.2%
Robbery is up 5.8%
Aggravated Assault is up 6.8%
Non-Fatal Shootings are up 21.2% (!!!)

You want to talk about how many police officers are being shot in 2016 compared to earlier years?

People are feeling less safe maybe because they are less safe. Newt didn't have a firm enough grasp on the subject before getting in front of the camera and John Oliver's staff gleefully cut up the clip for maximum effect. It was a funny segment on his show, but to cherry pick data after the fact to support the idea that Americans are mouth breathing idiots who don't know what's what is unfortunate. Do better.

[1] http://www.wsj.com/articles/ferguson-effect-is-real-and-hurts-minoritie…

[2] http://www.nytimes.com/interactive/2016/05/13/us/document-violent-crime…

Do better.

On that note, you forgot something.

To eric #6:

“Newt … basically admits that he, as a conservative politician, cares more about catering to people’s feelings than what’s factually true about the world.”

Then, that’s a strike against Newt.
But it would also be a strike against virtually all politicians, who would say the same thing if they were more honest and transparent.
The first priority of a politician is to get elected.
The second priority is to get re-elected.
The best way to do both is to tell the most voters what they WANT to hear, and to play to their FEELINGS.

It’s why Barack Obama was re-elected, and why Hillary Clinton may be elected.

It’s why the presidential race has seen virtually no discussion of the national debt, which, together with the present value of unfunded future liabilities for Social Security and Medicare, amounts to about $1,000,000 per American household.
People don't want to hear about it.

It’s why virtually all Dems are in favor of a national minimum wage - logic and equity and economics to the contrary. Sometimes they even admit it!
[“Economically, minimum wages may not make sense…Morally and socially and politically, they make every sense because it binds the community together and makes sure that parents can take care of their kids in a much more satisfactory way.”
- Governor Jerry Brown, during his signing of $15 minimum wage for Calif.]

Some things Newt might have asked of the CNN interviewer instead:
- ‘Yes, crime rates are dropping nationwide.
And gun sales are at all-time highs.
These are good, and perhaps even related things, yes?’

- ‘The FBI statistics show violent crimes nationwide dropping to about 1,200,000 in 2013. That’s good.
But let’s say that, hypothetically, we got violent crimes down to about 120, year after year. That would be fantastic, right? But what if in the latest hypothetical year the violent crimes went up to 150 (a 25% increase).
Would this be cause for national alarm and massive protests and news coverage? No. We’d still be in a fantastic position.
I say all this by way of introducing a question on
Black Lives Matter.
Killings of suspected felons of ALL RACES by police hit 461 in 2013, up from 367 in 1991 (a 25% increase). http://www.usatoday.com/story/news/nation/2014/11/11/police-killings-hu…

Do you think the Black Lives Matter movement runs on feelings (and agendas) more than facts?’

"These are good, and perhaps even related things, yes?’"
The first good, the second probably not, and no, if you mean cause and effect related, probably not. I don't doubt for a second you believe correlation is always causation, so chalk your statement here up to sheer stupidity rather than ignorance.

I'm not sure what an almost 3 year old article has to contribute to this, other than it comes from USA Today and the fact that they tend to use small words makes it easy for you to think you understand it. Did you notice that the article does not reference Black Lives Matter at all? There is no logical reason for your question to be linked to it - other than the fact that you are, again, trying to show your racism by implying BLM is somehow falsely asserting that their message is "Black Lives Matter MORE than those of anyone else?

@Narad #12

Do better.

On that note, you forgot something.

What's that?

It’s why the presidential race has seen virtually no discussion of the national debt, which, together with the present value of unfunded future liabilities for Social Security and Medicare, amounts to about $1,000,000 per American household.

I'm unsurprised that S.N. not only failed to cite his source (the earliest media invocation I can find is from 2011), but also screwed up the denominator.

I'm sure that he will determinedly reject his "benefits" from the repulsive schemes of Social Security and Medicare in a few years, though. (Assuming he's not already receiving them at age 61.)

What’s that?

The separate comment with footnotes hadn't appeared. It's a one-link limit here, as you may recall.

And gun sales are at all-time highs.

Whoops.* You probably should have thought harder about comment 2.

* Even more hilariously, the freaking Mises Institute takes a contrary position, but you're going to have to find that yourself.

@Narad #17

The separate comment with footnotes hadn’t appeared. It’s a one-link limit here, as you may recall.

I do recall. It is the whole reason I put source notes in separate posts.

When I first started writing it I was thinking I was going to cite more extensively. A lot has been written on the subject with one side labeling it 'The Ferguson Effect' with blame placed squarely on the shoulders of Black Lives Matter. The opposing viewpoint labels it with the much less racially charged term "The Viral Video Effect".

The math heads over at FiveThirtyEight even did a statistical analysis showing the phenomenon was real and outside the bounds of variability. While the cause may be admittedly debatable, the outright denial that more real live flesh and blood people are being killed or hurt by criminals recently is not.

I hold @Ethan in very high regard and this article was hard to accept. Not only did it pretend these victims don't exist, and back that up with data that "conveniently" ended right before the uptick started, but it also engaged in the type of Ad Hominem attack on Americans that @Ethan NEVER does. The cherry on top of the sundae is that it wasn't even original. This was a bit on comedy news program Last Week Tonight with John Oliver.

I love me some John Oliver, but @Ethan is not a comedian and this piece wasn't presented in that context. In truth I probably came of harsher than I meant to and would take back the first 2 lines if posts were editable, but this piece is disappointing on more than one level.

The separate comment with footnotes hadn’t appeared. It’s a one-link limit here, as you may recall.

I do recall. It is the whole reason I put source notes in separate posts.

Um, OK, but I'm not sure what the point of posting a partial comment while waiting for the references to appear is. In my case, it just led to irritation at trying to find them.

On the other hand, I can see a utility in compartmentalizing references in a subsequent comment per se, but why don't you just obfuscate them so they can be copied and pasted at the outset? I mean, this isn't difficult:

"http://www.wsj.com/articles/ferguson-effect-is-real-and-hurts-minoritie…

^ Not tossing out paywalled WSJ items would also be helpful.

Anyone wondering why the lefties take 1996 as their base year should look at this:

https://en.wikipedia.org/wiki/Violent_crime#/media/File:Violent_Crime_i…

And they are the ones who accuse their adversaries of cherry picking.
Many categories of violent crime are still 4 times higher than they were in 1960.

1996 corresponds with the peak influence of liberal ideology in relation to crime; since then more realistic policies have led to some improvement.
Perceptions are important, and they can be realistic. A look across the US's southern border at shows many countries that are effectively ruled by crime.
It's worth a lot of vigilance to prevent the US turning that way.

peter, how long did it take you to find a graph that exemplifies two of the worst ways to display data (raw numbers and multiple vertical axes)? I know it is from Wiikipedia and I'm not surprised, given their low quality of work with things statistical and mathematical. It's also interesting that murder and manslaughter are lumped together, given the legal distinction - but, since they are, if you look at the rates (which you should) over time: (data are available yearly: shown here every ten years, link below)

Year: Murder/Non-negligent manslaughter rate per 100000
1950 4.6
1960 5.1
1970 7.9
1980 10.2
1990 9.4
2000 5.5
2010 4.8

Data from
http://www.infoplease.com/ipa/A0873729.html

It's worth noting that the population of the United States has increased roughly 78% from 1960 (your cherry-picked year) to 2014. The overall crime rate (per 100,000) has increased about 57% - that doesn't quite match the distorted message you, sn, gingrich, or others send, whether by intentional dishonesty (sn, gingrich, many on the right) or simply because of blindly swallowing misleading representations of data. (I can't say for sure which category you are in.)

@Narad #21

Not tossing out paywalled WSJ items would also be helpful.

The WSJ is not paywalled from Google. To be more clear, here is how to read a WSJ article:

1 - Click on the link
2 - Highlight and copy the title of the article (in this case its 'Ferguson Effect Is Real and Hurts Minorities')
3 - Paste the article title into the Google search box and hit enter
4 - Click on the link in the search results and it will take you to the full article.

Aaaah, the "Ferguson effect" - the debunked idea that rising crime rates in select cities is somehow a harbinger the general situation. Pushed everywhere by people who would find another reason to scream "See, there is a boogeyman" if the FE hadn't been tossed to them.

Do violent gun deaths include suicide? Of the 33,000 gun deaths last year 60% were suicides on the other hand 43,000 deaths from second hand smoke were all homocides. 480,000 people died in the US last year from tobacco use. More money was spent on tobacco related health issues than any other single health issue. So no, I am not feeling the real need to hear about gun control when in a couple of years the death toll should be a cool half million dead from tobacco use.

"Now that may seem like a huge overreach, but in fact our entertainment culture really is pushing us in this direction."

No, it's the abject policies of a political class that has no idea what life is like in reality, a political class whose only "job" has been either politics or law.

It's the problem of the massive power of a political class (the legal profession) that is embedded in the idea that there is no "truth", just what "both sides" argue, and whoever "wins" the argument IS the "truth".

It's the problem of a political class (the mainstream media) who will hype up anything to present false balance in fear of the right wing politicians who LOVE their violence and in fear of money who want to keep things as they are, since if you're already at the top, you're most likely to do down if things change (even if you remain right in the top echelon).

The media isn't movies, it's the news media, purchased to keep things uncertain, scared and status quo.

"Here is a hint: We didn’t stop collecting crime data in 2013."

Here's a hint: no compiled statistic is up to the recent date, especially when culled across a huge landmass of a third of a billion people.

I realise that this isn't much of a hint for you, so I'll spell it out: THE RESULT IS NOT INVALIDATED AT ALL BY YOUR "HINT".

"Anyone wondering why the lefties take 1996 as their base year should look at this:"

Anyone wondering why Pete here is dismissive needs to look at the word "lefties".

IOW, this is a RWNJ who is shit scared because "dem niggas gonna git me and me wimmin!".

Because, after all, that's what HE'D do if he were in charge.

dean @23
I went for the graph covering the longest period and presented it (linked to it) uncropped.
Are we expected to believe the graph in the OP was just coincidentally cropped at 1996? So much for cherry picking.
On the whole, I think you're poorly placed to be passing judgement on other peoples' honesty.

You went for a graph that you could squint at and make believe there's a problem, because you're shit scared and want a "reason" for it.

"So much for the cherry picking", but that goes too for your first claims against "leftists". They got the earliest data that they had the same coverage and the same definition of the categories.

"On the whole, I think you’re poorly placed to be passing judgement on other peoples’ honesty."

Yes, but you're not going to be accepting your own dishonesty on this score, are you.

@Wow #27

Here’s a hint: no compiled statistic is up to the recent date, especially when culled across a huge landmass of a third of a billion people.

'Cuz they say two thousand zero zero party over,
Oops out of time
So tonight I'm gonna party like it's 1972

Do they not have networked computers where you live?

Pete, maybe you can show us where the data for people's perception of the rate of violent crime is pre-1996?

Getting your problem yet?

"" Here’s a hint: no compiled statistic is up to the recent date, especially when culled across a huge landmass of a third of a billion people."

‘Cuz they say two thousand zero zero party over,..."

And so you admit that your point is not only fatuous but misleading in that it indicates NO problem you allude to existing in the article.

After all, if the "best" you've got is a pop song, you must be admitting that your error is indefensible.

Or you're a blithering idiot, which means the same thing, but you won't be able to see it.

"Do they not have networked computers where you live?"

Yes, they do.

Are you capable of explaining why this question arose, or are you just blithering with a "LOOK! SQUIRREL!!!!" moment to hide your shame?

@Wow #25

In this day and age, networked computers have made statistics available on seemingly everything. Being that I had already included a link to very detailed crime statistics through the first part of 2016, I liked your breathless claim that compiled statistics on crime that were newer than 3 years old would be inconceivable.

I've been curious about clowns in Europe. Are you more like a mime? Or are you more like Ronald McDonald? Do you know how to make balloon animals?

and the misrepresentation of those statistics continues - why is the point of Ethan's article in dispute?

@dean #37

If you want to look at the statistics they are higher up in the thread, or if you want to pull them straight from the FBI they publish them right on their website. The violent crime rate nationally was up both in 2015 and so far in 2016, but it is the sleight of hand that is the point.

Alisyn Camerota assigned Newt the argument of denying FBI statistics. It was Alisyn that mockingly debunked the statistics by saying ‘Liberals use these numbers. They use this sort of magic math’. It was Alisyn that referred to the FBI as a liberal organization, all of which is of course ridiculous.

John Oliver’s staff got a hold of the segment, cut off the beginning where it made clear the topic was the appropriateness of the tone of the speech, and cut off the end for the same reason. A laugh track was inserted over the top to peer pressure viewers into seeing Newt’s comments as worth ridicule and everyone’s mind connected the dots making it Newt who had voiced those ridiculous claims.

As @eric pointed out in the other thread. I can show you exactly how the magic trick was done and most here will still eat the BS they were just spoon fed, exclaim how delicious it is, and ask for more please. As for me, I’ll calling it as the impressive bit of tradecraft that I see it to be. You can see it however you want.

@38: The statistics at the time are not national stats, unless there is a portion that didn't appear when I looked - I saw a table with categories, raw numbers for 2015 and for 2016. The percentage changes I calculated matched the numbers mentioned above, so I assume that that is the correct location.
Those are not national statistics.
The WSJ article is about the "Ferguson Effect" - an idea pushed by the some, but without any real support by the statistics or researchers.
* Johns Hopkins researchers found "fundamentally unclear and inconclusive evidence" of such an effect in Baltimore. After the arrests in Ferguson arrests in Baltimore fluctuated within levels expected based on recent history. After Freddie Grey died, arrests declined, murders did go up, but gun seizures in Baltimore had begun rising a year previous, in June. Shootings have gone back down.

* The data are equally confusing in Chicago. Murders are up (one of the cities often mentioned), as are gun seizures. The summary (good discussion at 5/38 a bit back) is that neither police officials nor researchers know what is going on for sure, but the Ferguson effect is not at the forefront for the reasons

* The line that "murders were up in 2015 and continue to be nationwide" may be true, but not in a very honest way: that rise is due to the data from 10 cities. Nationwide, levels are still far below the 1980s levels.

*There is an interesting theory from Richard Rosenfeld at U Missouri at St. Lois. He doesn't buy the "de-policing" form of the Ferguson effect (quote: There is no research I am aware of that suggests de-policing could have such a powerful effect on firearm violence -- except (maybe) if the police went on strike and stayed home." He favors the notion that

...longstanding grievances with the police in minority communities are activated by controversial and heavily publicized incidents of police use of force, resulting in more killings as community members settle grievances or respond to crimes without recourse to the police.

None of that jibes with New York: stop and frisk has undergone a dramatic decline, but the murder rate has also declined: Q1 2016 had the fewest shootings and murders in the city's history.

So no, there was nothing wrong with my comment at 37. Gingrich, and others, are falsely trying to conflate statistics from a few regions and convince people they indicate a national trend, when nothing could be further from the truth.

@dean #39

The WSJ article is about the “Ferguson Effect”

I don't know about your unsupported attack on the Ferguson Effect that has nothing to do with anything, but the 2015 violent crime statistics I cited from that article were taken straight from the FBI 2015 Uniform Crime Report. They are national statistics and they are up.

Of course the FBI is a liberal organization and they have magic math theoriticians so who knows what's really going on, or so Newt said when he was lying about universal truths.

The link that took me to WSJ took me to an article on the FE. (It isn't just me that believes it is bunk, most researchers do as well.) There were no statistics to be seen. I will try the link later to make sure I didn't change a page once I was there.

@dean #41

Here is the direct link to the FBI:

FBI 2015 Uniform Crime Report

A newt is a semiaquatic amphibian of the family Salamandridae, although not all aquatic salamanders are considered newts. Newts are classified as a part of the salamandrid subfamily Pleurodelinae, and can be found in North America ............
Newts are semi-aquatic amphibians that look a bit like a combination of a frog and a lizard.
Hmmmm .....

"In this day and age, networked computers have made statistics available on seemingly everything."

Pointless verbiage, get to the point, idiot.

"Being that I had already included a link to very detailed crime statistics through the first part of 2016,"

But fuck all about how people THINK the prior years' crime statistics turned out, your "point" is entirely pointless still.

Hell, even on the basis that a 9% rise would STILL leave "people" claiming MORE THAN TWICE the crime rate actually in place, your "point" is entirely specious.

But if peiople "think" it's more than 71 per thousand households, they're EVEN WORSE chicken littles.

It's "amusing" that you "think" still that you have a point when it's failed on every salient metric.

Ethan, I wished you would have included John Lott 's work in this article and more precisely his work on Gun control hysteria from the left wind libs and based it on crime stats. He wrote a great book titled "More Guns, Less Crime" http://johnrlott.blogspot.com/

Lott's work where he was found to have faked data? The results he claimed were based on his own survey, the one where he couldn't produced the data when asked ("It was lost"), couldn't show the questions he used? ("I don't know where they are.") and could not name any of the "researchers" who supposedly helped him? The work where his "model" shows rural areas to be more dangerous than inner cities?

Lott's work has been repeatedly shown to be garbage.

dean,
do you actually research the lies you spew?
Here is the data you falsely claim he doesn't have.:
http://crimeresearch.org/data/

I know much more than you do rt: I've covered Lott's failed work in graduate classes under "How not to do modeling."

For the data issue:
a) Lott claimed that "national surveys" supported his claim that 95% of the time people used guns defensively they had to brandish the weapon to break off an attack". He said the polls were from the LA Times, Gallup, and Peter Hart Research Associates. When these surveys couldn't be found, he claimed he used data from his own survey, done with help of research assistants. Asked for the data so that others could examine it, Lott claimed it didn't exist because it was lost in a computer crash . The great unlikelihood of important data being stored on a single computer was immediately noted. Neither could Lott supply the name of any research associate, nor the survey questions, or even evidence of phone records, to show the survey actually existed. This is not to say that a "defender" of Lott didn't appear:

(After criticisms began) someone named “Mary Rosh” emerged as one of his most vocal defenders online. “Rosh” wrote positive reviews of Lott’s book on Amazon, and she showed up in comment forums on various websites to describe Lott as a “meticulous researcher” and “the best professor I ever had.”

But: Rosh was Lott himself. Julian Sanchez, now a fellow at the Cato Institute,

compared the I.P. address on Rosh’s comment to the one on an e-mail Lott had sent me from his home. They were the same.” When Sanchez confronted Lott with this evidence, Lott confessed that “the MaRyRoSh pen name account was created years ago for an account for my children, using the first two letters of the names of my four sons.

The data hasn't been seen. So: no evidence it existed, no evidence of a survey - not much evidence in favor of it existing.

It isn't just me: some points from criminologists and statisticians around the country.

Lott argued there were more guns from 1977 to 1992. He did this by picking two different surveys and “adjusting” for the differences (not a statistical adjustment, an arbitrary adjustment of his choosing). This assertion, however, contradicts 86 surveys on gun ownership, as well as sales records and registration
* Lott replied that even if there weren’t more households with guns, more people were carrying guns in public due to changes in the law. Criminologists pointed out that this could not be responsible for the decrease in crime: records and surveys showed 5-11% of the public were already carrying guns for self defense, and a 1% increase in the number (which was the actual increase) could not explain the corresponding decrease in crime
* Lott responded that the “new” 1% lived in locations in which the risk of crime was higher than others, and that was the cause of the decrease. Several studies showed that zip codes with the highest violent crime rates had no, or smallest in their region, increase in new permits

Lott relied heavily on Dade county to support his argument. Records on 100,000 (one hundred thousand) violent crimes over the time period he claimed to have studied revealed 12 (twelve) cases where a person with a concealed weapon was involved. That is 0.012%. Gary Kleck, a criminologist at FSU, stated (for this and other reasons touched on later) that the crime decrease observed there was not due to the increase in carry but other reasons which Lott had neither considered nor controlled for in his model.

More than that: the model, despite the repeated refinements Lott claims to have made, is just bad. When we develop a model we look for (or should look for) a model that shows statistical significance for variables that can reasonably be considered relevant - in large data sets significance can appear for any data, simply as a by-product of sample size. The model should represent the existing data well: one way to check this (the simplest way) is to split the data into two groups, at random, build it with one group, and compare its predictions with the observed values based on the second group. Predictions from the model should also be compared to newly collected data - a good model will compare reasonably well.
The model should be reasonably robust (intuitive meaning: model significance and predictions should not change drastically when there are changes to a portion of the data). Lott's model fails all of these.

1) It indicated rural areas would be more dangerous than inner city areas - in contradiction of years of local, state, and federal data
2) His model said that if unemployment were INCREASED and the number of black women middle-aged and older DECREASED, the result would be a dramatic decrease in the homicide rate. Even more odd,

a decrease of 1 percentage point in the percentage of the population that is black, female, and aged 40 to 49 is associated with a 59% DECREASE in homicide and a 74% INCREASE in rape

3) The model indicated concealed-carry laws were associated with a decrease in murder and rape and with an increase in the rate of property crime. He asserted that this was due to criminals who would usually commit predatory crimes switching to property crimes. That generated (essentially) laughter from criminologists.
4) Crime rates are cyclical - they rise and fall over time, changing with patterns of gang membership, drug availability and consumption, availability of illegal guns, the economy, and more. Typically this is modeled with (surprise) time series methods. Lott did not do that: he included a linear time trend for the crime rate, which had the effect of introducing the (false) assumption that crime always increases at the same constant rate.
5) Lack of robustness. ALL of the positive effects predicted by Lott's model are due the influence of data from Florida: Black and Nagin (criminologists) identified the high influence level of Florida (there is a specific statistical meaning for "influence" in this type of modeling) and found that, when Florida was removed, the positive effects predicted by increasing gun availability vanished. What made Florida key? They determined that during the time period covered by Lott's data Florida's crime rate was highly volatile, changing more rapidly and more drastically than any other state.
6) Bad predictive power. It is easy to make a model that looks impressive, but if it can't make predictions that match real data it is worthless. If Lott's model were good, it would be able to make reliable predictions for data outside that from which it was constructed.

Researchers Ian Ayres, from Yale Law School, and John Donohue, from Stanford Law School, did just this, and examined 14 additional jurisdictions between 1992 and 1996 that adopted concealed carry laws. Using Lott’s own model, they found that these jurisdictions were associated with more crime in all crime categories. In other words, “More Guns, More Crime.” Ayres and Donohue conclude with the rather damning paragraph, “Those who were swayed by the statistical evidence previously offered by Lott and Mustard to believe the more guns, less crime hypothesis should now be more strongly inclined to accept the even stronger statistical evidence suggesting the crime inducing effect of shall issue laws.

Not surprisingly, Lott tried to address this.

John Lott, along with two other young researchers, Florenz Plassmann and John Whitely, wrote a reply to Ayres and Donohue attempting to confirm the “more guns, less crime” hypothesis. However, when Ayres and Donohue examined Lott’s reply, they discovered numerous coding errors and empty cells that, when corrected, showed that RTC laws did not reduce crime and in some categories even increased it. In a latter email exchange with Tim Lambert, Plassmann even admitted that correcting the coding errors caused his paper’s conclusions to evaporate. Eventually, Lott removed his name from the final paper, citing disagreements over edits (which turned out to be a conflict over a single word) that had been made to Ayres and Donohue’s paper.

There was one other issue. There was a great deal of discussion about how much the missing values in Lott's data influenced results. The analysis strongly indicated they did: Lott claimed they did not. To support his case Lott posted a table which he claimed had the corrected data and that results were still significant. This seemed to contradict Ayres and Donahue. There was confusion until it was discovered that Lott had changed the rules for the analysis without revealing that change.

In the original analysis Ayres and Donahue assumed clustered standard errors. In his rebuttal Lott removed that requirement, and that was enough to make the results significant.

Finally, the National Research Council convened a panel to examine literature and research. On right to carry laws the panel attempted to replicate, for purposes of investigation, Lott's work. They noted the problems mentioned above, and more, and 15 of the 16 panel members concluded that based on the existing work, there was no support for the the claims that RTC had either a beneficial or a detrimental impact on crime rates.

aight, deano, give me a day or so to respond. I have had 1 hour sleep in the last 29 ( I had to work overtime) and will be refreshed tomorrow.
Please refresh my memory on your academic degree of prowess. It was some sort of numbers/data modeling degree wasn't it?

Teabaggie, in what way does Dean's academic degree (which you yourself do not posses in the least) change the facts of the case that can be understood WITHOUT a degree?

After all, if even Einstein had forgotten to take his coat with him, even the smallest and dumbest child could observe this fact, not only the equivalently educated.

@ Retard # 51
QUOTE "Teabaggie, in what way does Dean’s academic degree (which you yourself do not posses in the least) change the facts of the case that can be understood WITHOUT a degree"
Well, tardo, his "Degree" in the past would give him some semblance of credibility. However, nowadays, When Niggers are given degrees simply because they are ...well Niggers AKA a disenfranchised segment of society that the lefttard liberals have deemed the "New Vitim" class for political gain..
So Fuck Off Peter Breath..
Crawl back under your pathetic slimy hole you crawled out of. Of which mankind is now worse off from your pathetic contribution.

So Retard,

Clean up on aisle five.
Ethan, if you're going to consider polls/votes to ban commenters, consider #52. RM's bigotry is frankly more offensive than anything SN has to say.

Wow. so much for there being any chance rt would be able to understand anything I posted about Lott - #52 demonstrates he has no concept of anything academic. Or decent.