86 that Rand 420 study: 5th worst of 2011

[RainDog]

from Scientific American
http://www.scientificamerican.com/article.cfm?id=top-science-journal

Doh! Top Science Journal Retractions of 2011

#5: Los Angeles marijuana dispensaries lead to drop in crime.

Keep smoking. The RAND Corporation retracted its own report in October after realizing its sloppy data collection.

Crime data compiled from neighborhoods with these highly contentious medical marijuana dispensaries supposedly revealed slightly lower crime rates. The authors attributed this decline not to marijuana itself but rather the presence of security cameras and guards in and around the dispensaries, having a positive effect on the neighborhood. [The History of 8 Hallucinogens]

The L.A. city attorney's office was incensed with the report, having argued the opposite — that the dispensaries breed crime. The city's lawyers soon found critical flaws in RAND's data collection, largely stemming from RAND's reliance on data from CrimeReports.com, which did not include data from the L.A. Police Department. RAND blamed itself for the error, not CrimeReports.com, which had made no claims of having a complete set of data, and, in fact, didn't even know about the study.

This was posted here before. Just wanted to make sure it got notice.

[killbotfactory]

1. What evidence does the LA attorney's office provide to prove they "breed" crime? nt

[killbotfactory]

2. Oh, and also, in what ways would data from the LAPD change their report? nt

[RainDog]

9. we don't know. The LAPD doesn't want the study to be true

so they went after it.

Rand should do another study with data from lots of diff. cities in diff. states - there are enough places now that could provide a better sample and would dilute the desire to disprove such an idea if it is true.

the study claimed the reason crime was down was b/c of enhanced security, not because cannabis stopped crime.

the LAPD doesn't want to find anything good about dispensaries. the study is effectively void.

proves nothing one way or another.

[tridim]

4. Because marijuana makes you smile and listen to jazz.

Apparently two of the worst things in the whole world.

It breeds "crime" because it is not yet federally legal.

[RainDog]

5. the point is that the Rand study assumed it had data that it didn't

it doesn't mean the study might still be true but they hired out data collection and didn't know what data they were including or excluding.

iow, they should do another study and use better data. it still might be true, but not in the way it was presented.

or it might not be true.

iow, there's no valid study to claim anything.

[REP]

3. I'm not sure I trust the LAPD/LA Atty about much of anything

One of the most historically corrupt police departments and city attorney offices in the country.

[RainDog]

6. well, yeah, police depts massage data all the time

to meet numbers, etc.

but the study thought it was looking at data from the LAPD and it wasn't. that's what makes it a bad study.

[REP]

7. Not saying it was a good study - just think LAPD/atty have an agenda

and they have a record of not being all too trustworthy, flawed study or not.

[RainDog]

13. no doubt. they were the ones that looked into the study to question it

but that doesn't mean anything regarding the study itself - the problem was a claim made in the study that relied upon data it thought was included. the data wasn't included b/c the company that gave them the info didn't include it.

that's all.

but, for the sake of accuracy, it's good to know that this study is not considered valid b/c the data is flawed.

[REP]

14. I agree

And I wasn't disagreeing with any of your points - just grumblingly complaining about LAPD etc. You mean you can't see me nodding my head??

A good study with good data would be interesting.

[quakerboy]

8. I am willing to bet that crime increased around the dispensaries

Given that marijuana is a federal crime, it would pretty much have to do so, inherently.. I bet that given the wave of federal raids, the crime stats in those neighborhoods have significantly spiked.

It would be nice if this "article" had a link to any of the things it asserts

[RainDog]

11. your bet doesn't look at the issue either

if you googled Rand study and dispensaries, I bet you could find it.

oh wait.. look... less than a minute and here it is!

http://www.rand.org/pubs/technical_reports/TR987.html

[quakerboy]

16. It wasnt intended to

A well written roundup article should have included the source links to each of these things it referred to.

It will be interesting to see what their new review comes up with. But I am not certain that I will trust it any more than the original. As others have pointed out, while there may be an issue with the data used, I am not convinced that using LAPD data will bring us closer to the truth. LAPD data will reflect crime being where ever LAPD chooses to enforce the laws. Which is a flawed metric.

[RainDog]

17. people report crimes

whether the law is enforced or not.

crime reports come from people who make complaints.

[cthulu2016]

10. A fraudulent study is usually a counter-indicator. A flawed study is not.

When someone fakes a study it tends to mean that they could not get the desired result honestly and the study will tend to have false conclusions.

When a study is flawed in the way this one was it doesn't provide clues to the real state of things. It just lets you know that the study is not reliable. Knocks you back to square one, as if no study had been done at all.

[RainDog]

12. exactly n/t

[bemildred]

15. Dumb studies produce dumb results. nt