Of Questioning Data Quality and Government

Oh my guilty pleasure, I'm a lurker on the comment sections in various Internet corners. I enjoy cunning replies, snappy comebacks, and clever insults over various debated topics. You never really know when you’ll find a little gem but it’s easy to surmise topics that are more likely to get folks provoked.

Why do I make this disclosure? It's because of the Wall Street Journal What Is the State of the US Gun Business? post stating household gun ownership had decreased 20% compared to the 1970s according to IBISWorld estimate. Immediately I scrolled to the comments section expecting a healthy issue examination but was surprised that most comments were about the data. Eek, embarrassing for the author!

Foul says the commenters Most of the commenters questioned how the data was correct. Commenters noted that more guns sales are rising, especially to women, and more people are applying for concealed carry permits. They have a point, the survey results do not jibe. Gallup agreed there was a drop, but it was closer to 7%. Those survey results are far apart as a strategy meeting with Willy Wonka, Yosemite Sam, and Dilbert.

Various points of view are offered about the difference. The most obvious target is how the data is collected. Some commenters thought how the researchers worded the question was the issue. For instance, Gallup asked if there were any guns on the property (including the garage) thus their results were higher than those questioned just about the house. Apparently, people don't consider the garage a part of the house or were asked about a gun instead of a firearm and they don't consider a rifle a gun. Another possibility is the person forgot such as cases where a hunting rifle was inherited or maybe the household member asked didn’t know.

Others suggested the differences represent, as Pew Research discusses, how data was collected, suggesting a difference between phone, web, and door-to-door surveys. Can you imagine a door-to-door surveyor asking if you have a gun? Don't you think some might wonder if they were being cased? I’m sure it was a more formal methodology than a stranger landing on someone’s doorstep, or at least I hope so.

Paranoid or enlightened viewpoint? Some comments are a stinging indictment revealing that gun owners are not responding truthfully because of the potential data misuse. One commenter said, "Make the government go through their records to figure it out." That is not paranoia -- it’s an enlightened viewpoint! People are recognizing that once data leaves your lips and enters a computer it has its own life, and it’s a long one.

This is the most curious aspect because it means what data people have long known: It is not that hard to combine data. I can’t help but reflect on the impact that whistleblower Ed Snowden, the Patriot act, and the media spotlight have had on a rising national suspicion about data usage. Even if the survey company is promising anonymity and only interested in aggregated results, they still have the transactional data.

I’m not alone Pew Research found that of the 87% of Americanswho are aware of government surveillance programs, 34% had taken steps to shield their information from the government. Other interesting results include how people are changing their usage patterns and have taken steps to increase their privacy. However, I have to ask, are they answering truthfully? Like the gun owners, it seems silly to reveal mistrust.