Gun Control Reform Activism

named Gary Kleck. His 'study' is so deeply flawed that any reputable statistician will laugh until they collapse from exhaustion when they look at it.

What Kleck did:

Using standard polling procedure he randomly dialed 5000 phone numbers and asked a series of questions that supposedly qualified the responses.

What Kleck did wrong:

At the time of his 'study' only 35% of households had guns and could possibly answer yes to the DGU question. So immediately 65% of the sample was useless. Based on the answers he got he extrapolated the results to all 300 million Americans which again only 35% could actually claim a DGU.

Second he used data that could not be corroborated through empirical measures. No police reports, no secondary witnesses, NOTHING to lend credence to the claims of the respondents.

Third he asked a highly loaded question to an audience with a bias to responding 'yes'. The possibility of getting false positives is extremely high in these situations as people who have and reallllllly like guns want to justify the importance of having them. If I wanted to get a lot of false negatives I'd ask if the respondent routinely watches porn and just accept their response without looking a their browser history.

Fourth he failed to correct for the 'telescoping effect'. This effect tends to inflate results because it uses information from outside the time window of the 'study', i.e. was the DGU in the 12 month window or was it 15 months ago?

So Kleck's 'study' wasn't really a study. It was more like a poll and a very poorly conducted poll at that.