When 85% of your study’s data comes from 6% of the states, two of which are known as being anti-gun, you have a poor study and an even worse credibility score.

Critique

• 85% of reporting law enforcement agencies come from just three states (6% of states).
• 51% of reporting law enforcement agencies come from just California.
• A mere 1,053 agencies participated (there are over 3,000 counties in the U.S. – multiply by the number of city agencies).

Background

First, let’s take NBC to task. The Trace is openly backed by Michael Bloomberg, a billionaire politician with a messianic quest to enact gun control laws. For a news organization to collaborate with an established front group calls NBC’s objectivity into question, though many people doubted their objectivity to begin with.

“Every news organization has propagandist tendencies, except for MSNBC, which is a propaganda organization with news tendencies.” — Guy Smith

What the Trace attempted was to contact law enforcement agencies and gather data on stolen guns, as well as some data on recovered crime guns. The alleged goal was to better understand the impact stolen guns have on the overall crime-gun situation.

The disturbing details

I wish I could be more generous, but their data is not remotely close to criminology standards, or even the low standards required of freshman quantitative research students.

Their basic approach was to ask for data, toss it into a data set, perform some fundamental data cleansing, and then report. The problem with this approach is that it has no goal. No findings. No hypothesis to test. Without such, the outcome is meaningless.

Worse yet is that they published with what can charitably be called “research data”. That more than half of all reporting agencies came from one state – a state that funds a notoriously anti-gun researcher and has the strictest gun control laws in the country – kinda skews the results. California is a model for nothing else in the United States, and that it contributes 51% of all reporting agencies makes the entire data set valueless. The heat map above shows how skewed their data gathering is.

Worse still is that nearly all (85.1%) of reporting agencies come from just three states, with Illinois (another state with a history strict gun control) being in the list.

This leads to some rather disturbing anomalies. Take the spread between California and Texas. The table to the right shows the number of counties in both states, the number of agencies (county, city, state) that contributed to Trace, and the county/agency ratio.

Anyone who thinks this is a viable data set must work for NBC.

Though there is some breathless “reporting” in the article, there appears to be no attempt to discover how stolen guns migrate into the criminal network, or the path they take from entry to use in crime. Nor does the report examine the ratio of guns stolen from civilians and from law enforcement ( the latter group has a surprisingly high number of news reports of their guns being heisted).

The take-away

There is none. The only mention this work should receive is that it is meaningless in understanding how stolen guns figure into crime. Log it not as debunked, but dismissed.