Gun Rights: Part 3: Gun Regulation: Will Reasonable Gun Control Save Lives?
WHAT HAVE WE ESTABLISHED SO FAR?
 THAT THERE IS A VERY STRONG STATISTICAL CORRELATION between the Rate of Deaths from all Causes in a given geographic area and the Rate of Gun Ownership in that same locality.
 THAT THAT THERE IS NO STATISTICAL CORRELATION between the Rate of Violent Crime in a given geographic area and the Rate of Gun Ownership in that same locality.
 THAT THAT THERE IS A POTENTIAL STATISTICAL CORRELATION between the Rate of Violent Crime and the Rate of Gun Ownership, when combined with other related factors. Further, the statistics seem to indicate that the relationship with gun ownership is negative, meaning more guns, less violent crime, but ONLY when certain other factors are present in the right quantities.
Since I have mentioned total deaths and gun control advocates seem to want to minimize this aspect of gun ownership, I might as well show you some statistics regarding that, they might surprise you.
Articles on Sensible Gun Regulations That the NRA Hates
 As a result of the several mass shootings in Virginia and the killing of the WDBJ news crew, Virginia has stopped accepting recipracle concealedcarry permits from States with looser permit requirements than Virginia has. Basically it denies recognition of States who allow felons, mentally ill, and those charged with domestic violence or are under a restraining order to carry concealed.  http://money.cnn.com/2015/12/23/news/virginiaconcealedcarrygunpermits/index.html?iid=EL
CAUSE OF DEATH BY GUN  NATIONAL
CAUSE
 AMOUNT (2007)
 RATE PER 100,000
 % OF TOTAL


SUICIDE
 17,348
 5.8
 56%

HOMICIDE
 12,129
 4.0
 39%

ACCIDENT
 721
 .2
 2%

LEGAL INTERVENTION
 315
 .1
 2%

OTHER
 256
 .1
 1%

TOTAL
 30,769
 9.9

We see that homicides account for only 39% of deaths while suicides amount to 56%, over half of all gun deaths! These are, however, national statistics; of more interest would be similar statistics by state, over time. Why I feel this way is, I suspect, the reasons for death by gun differ substantially by state and that difference could have a real impact on what the statistics tell us. But that will have to wait for a later time, a later hub.
Right now, I need establish the link between the degree a state regulates gun ownership, or the lack thereof, and the rate of gun ownership in each state. So, let me give you the answer up front and if you have a mind, you can continue reading to determine how I arrived at this answer.
The bottom line is this;
There is a statistically significant correlation between the rate of legal gun ownership and a combination of:
 The strength of state regulation of guns
 The population density of a given state
 The political makeup (more Republican or more Democratic)
On the scales used, each point increase in the
 The strength of regulations decreases the rate of legal gun ownership by about 0.2 percentage points; the range of ratings goes from 1 to 87 with 87 being the strongest regulation.
 The population density decreases the rate of legal gun ownership by about 0.02 percentage points, where the range of densities run from 1 to 1200.
 Actually, 1/10th of a point increase in Democratic representation will decrease the rate of legal gun ownership by around 2 percentage points. The range here goes from 0.13 to 0.90
IF A = B, and B = C, then A must = C
A PIECE OF TRANSITIVE LOGIC THAT APPLIES IN ALL but the most esoteric situations; and I will be using it in a minute. As I just mentioned, in the last hub we basically establish A = B or Rate of Gun Ownership is proportional to Rate of Death from all causes. Now I need to establish B = C, which will be the Rate of Gun Ownership to the Strength of Gun Control Regulation.
One of the biggest debates raging across the country right now, because of such recent massmurders as Columbine, Gabrielle Giffords, Aurora Theater, and Newtown, is what degree of state regulation produces the best results. Does more gun control lead to less deaths by gun from all causes? What is the distribution of the amount of regulation between the various states? How does that distribution impact results?
The statistics to show this relationship are fairly simple. What isn't necessarily simple is coming up with useful numbers is the Strength of State Gun Control Regulations. I was fortunate though and came across a study by "The Open Society Institute's Center on Crime, Community, and Culture's" report, Gun Control in the United States: A Comparative Study of State Firearm Laws. It established 30 different criteria with which to rate each state by and then ordered the states by the resulting totals.
The criteria were grouped into six categories:
 Registration of Firearms with 9 subcriteria
 Safety Training with 1 subcriteria
 Regulation of Firearms with 11 subcriteria
 Safety and Storage with 2 subcriteria
 Owner Licensing with 5 subcriteria
 Litigation and Preemption with 2 subcriteria
The results of their work is presented in column 'g' of Table 2 in Part 1. The negative values represent states whose laws detract from the minimum standards set by the federal government.
Negative numbers in arrays like this often present problems in analysis, so I got rid of them by adding 10 to each result when I actually used the data in my calculations. Another issue is that a set of numbers which relate to each other in an ordinal fashion (a larger number having a degree of significance greater than the number below it) is statistically useful in only certain respects, but not in others.
Where ordinal numbers are useful is in comparing "rankings" between two sets of ranks, in our case comparing the ranking of Degree of Gun Regulations with the ranking of % Gun Ownership. The statistical method we will use here is the Pearson's Rank Order Correlation. Where rankings are not useful is in calculating equations such as I did in Parts 1 and 2. The reason is, while a rank of 20 may by higher than a rank of 10, it isn't necessary "twice" as high; you can't assume the 2 to 1 relationship the rankings suggest. Regression analysis requires this kind of relationship exist before its output is valid.
Consequently, I must do other manipulation of the Comparative's Study's results to arrange them in such a way that 2 does mean it is twice has "important" as 1. Fortunately, I discovered a took when working in the Air Force, one they actually bought and used for decision making, which allows a user, or group of users, to convert "subjective" judgments into "objective" numeric relationships. I used this took here to "objectify" the Study's results before running my regression analyses.
ANALYSIS 1  PEARSON PRODUCTMOMENT CORRELATION
THE PEARSON PRODUCTMOMENT CORRELATION compares pairs of related numbers in a special way that is statistically relevant. Consider the thee sets of numbers below.
SERIES 1
 Ranking A
 Ranking B


1
 7
 
2
 8
 
3
 9
 
4
 10
 
Pearson's Correlation = 1
 
SERIES 2
 1
 10

2
 9
 
3
 8
 
4
 7
 
Pearson's Correlation = 1
 
SERIES 3
 
1
 9
 
2
 7
 
3
 10
 
4
 8
 
Pearson's Correlation = 0

So you see we have three results, 1, 1, and 0, one for each series. Series 1 is obviously positively correlated, and the 1.0 result says it is perfectly correlated. The second series is the same, except in a negative direction. The last series has a Pearson Correlation of zero, which of course means the two lists of numbers have no relation to each other, which, on inspection, you can easily tell is true.
Well. what I have done with our data is exactly the same thing. Rank A is the list of % Gun Ownership by state and the Rank B is the list of scores for state gun regulations. Based on the above information:
 If they were perfectly correlated, with a score of 1, that would mean the lowest ranking gun regulation score would coincide with the lowest percentage of gun ownership; the second lowest regulation with second lowest ownership; and so on. (There are a few more caveats, but I will skip those because they don't change the basic idea.)
 A minus one score means the lowest of one matches with the highest of the other, the second lowest with the second highest, etc.
 A zero score would mean the rank of A is completely independent of the rank of B.
The probabilities surrounding the Pearson Correlation suggest that for a level of significance of .05, the correlation needs to be greater than 0.28. In fact, when I run the numbers with Columns 'b' and 'g' of Table 2 in Part 1 I end up with a correlation of 0.75; a very significant result which strongly suggests a negative correlation between gun regulation and the % of Gun Ownership by state, meaning the more Gun Regulation there is, the lower the % Ownership of Guns in a given state. This, of course, is common sense.
REGRESSION ANALYSIS RESULTS
DON'T WORRY, I AM NOT GOING to bore you or make your eyes hurt with more charts, you have already seen how that works in Parts 1 and 2. I will just give you the summary results of my latest findings. As I did before, I looked at several likely independent variables, tried them out in different arrangements that made common sense and settled on one which gave reasonable results. Can I find a better model? Probably, but that is not the point, I am just trying to show that relationships do exist in the first place.
Updated 8/3/14 with new information: The independent variables I ended up in my model to predict % Ownership of Guns in a state are 1) Political Makeup, 2) Strength of Regulation, and 3) Population Density per square mile.
 The Political Makeup, in this case, is the % of Democrats that make up the each State's Legislature.
 The Strength of the Regulation is the result of my application of the results from the Comparative Study (column h, Table 2, Part 1) to the Analytical Hierarchy Process which converts ordinal numbers into hierarchical numbers that can be used in regressions.
 The Population Density is just the state's population density.
When I run all 50 States worth of data through the Excel regression program I get the following:
 An Adjusted R^{2} of: 71% (a reasonable result)
 A Significance F of 3.49 E13 (which is less than the .05 threshold)
 An Intercept (where the trend line crosses the YAxis) pvalue of almost zero (great)
 A Political Makeup pvalue of .009 (good)
 A Strength of Regulation pvalue of .01 (OK)
 A Population Density pvalue of .0001 (great)
So, all of the statistics about our model say we have a reasonably good model for predicting the % of Gun Ownership in a state based on those three variables from a State. The actual formula is:
% Gun Ownership = 55.30177  20.4867 * Political Makeup  .18676 * Regulation Strength  .02083 * Population Density
The variables can take on the following values:
Political Makeup: .13333  .9017 (Low is Democratic)
Regulation Strength: 1  87 (Lowest is less than Federal guidelines)
Population Density: 1.26  1205
So, what does all of this say? It says 1) the more Republican a state is, the higher the rate of gun ownership, 2) the less regulation on guns, the higher the rate of gun ownership, and 3) the lower the population density, the higher the rate of gun ownership.
This result says nothing about the relationship between gun regulation and the rate of deaths due to guns or the rate of violent crime. It is only showing there is a definite link between gun regulation and the rate of gun ownership.
THE LOGIC OF IT ALL
IN AN EARLIER SECTION I offered that if A = B and B = C, then A = C; a standard transitive logical statement. But first, a word about the word 'transitive". Transitive, in a sense, means that whatever word is substituted for the "equal" sign has exactly the same relationship in each case. Words like "greater than", "implies", "is a subset of", are all transitive. On the other hand, words like "love", "is the son of", "killed" are all intransitive and the logical statement breaks down. In our case, we are using the word "implies".
For example, an "intransitive case" would be IF A "loves" B and B "loves" C, then it necessarily does not follow that A "loves" C; by observation A may or may not "love" C. In the "transitive case", however, it does necessarily follow that IF A "implies" B and B "implies" C, THEN A must "imply" C.
So, what we are concluding in Part 3 is
 "an increase in the strength of gun regulation" (A) implies "a decrease in the rate of gun ownership" (B)
 and from Part 1 we determined that, "an decrease in the rate of gun ownership" (B) implies "an decrease in total deaths by gun" (C),
 we can properly assert that "an increase in the strength of gun regulation" (A) will result in "a decrease in total deaths by gun" (C).
This is exactly what we are still seeing after the passage of the Brady Handgun Violence Prevention and Violent Crime and Law Enforcement Acts, in 1994 with the subsequent longterm decline violent crime and deaths from guns. It must be noted, without comment, that the rate of decline decreased after 2001.
In Part 4, I will pick up this theme of A = B, B = C, A = C and develop it much further as it is central to my thesis that sensible, nationallyadopted, gun control laws will make a real, provable, positive difference in the lives of Americans and bring our death and violent crime rates which result from gun ownership down to be more in line with other developed countries.
A SIDE NOTE ON THE EBAY AD BELOW
AS A GOOD MEYERS BRIGG INTP that I am, I had to think further about a Tshirt being offered for sale for awhile in the EBAY ad space below. On it, in part, were the words,
"If Guns Kill People, Then Pencils Misspell Words ..."
This actually makes quite a lot sense on the face of it, until, that is, you dig a little deeper. A couple of things occurred to me for you to ponder:
 If guns had never been invented, there would be a lot more people alive today; but, if pencils (or similar writing instruments) had never been invented, you wouldn't be reading this or anything else, for that matter
 If a pencil falls off a table, since it was not in somebody's hand, a word did not get misspelled; if, however, a gun fell of a table, someone might still be killed even though nobody was holding it.
REFORMULATED QUESTION TO MAKE IT MORE NEUTRAL
DO YOU THINK MORE SENSIBLE GUN REGULATION WILL ...
I CHANGED THE FORMULATION of the question in the above poll to make it more neutral; the previous one sounded somewhat leading. I hope the four people who voted in the earlier poll (1 Yes, 2No, and 1Not Sure) will vote in this one again.