HomeLatest ThreadsGreatest ThreadsForums & GroupsMy SubscriptionsMy Posts
DU Home » Latest Threads » Forums & Groups » Main » General Discussion (Forum) » My challenge to DU: Cite ... » Reply #38

Response to Bjorn Against (Original post)

Fri May 17, 2013, 04:04 PM

38. Bayes Rule is useful here.

The fact that about 25% of scrutinized applications were Tea Party ones doesn't mean much by itself. You have to also know what fraction of the original applications were Tea Party ones, or similarly, what fraction of Tea Party applications were approved without review, compared to how many other ones.

As I wrote in another thread, suppose you found out that about 25% of drivers being pulled over in a state are black. Would this be evidence that police are not engaging in racial profiling? No-- you'd have to know what fraction of the population is black. If it turns out that blacks only make up 5% of the population, then you have a problem.

The inspector general's report says that 70% of applications overall pass without review. However, it appears that almost no Tea Party applications made it through without review (they found none in a sample of approved applications). Given that, and the fact that Tea Party terms were on the BOLO list, means that the Tea Party applications were in fact given special, negative treatment.

Does this mean that only Tea Party applications got negative treatment? No, it's possible other terms were also (inappropriately) added to the BOLO list. I would be curious as to what those were as well. But the inspector general report makes it clear that using the organization's name and policy positions as criteria is wrong, and the IRS was wrong to do so.

Reply to this post

Back to OP Alert abuse Link to post in-thread

Always highlight: 10 newest replies | Replies posted after I mark a forum
Replies to this discussion thread
Arrow 40 replies Author Time Post
Bjorn Against May 2013 OP
The Magistrate May 2013 #1
Vinnie From Indy May 2013 #2
Cha May 2013 #3
Bjorn Against May 2013 #4
Cha May 2013 #6
Control-Z May 2013 #10
Bjorn Against May 2013 #5
Cha May 2013 #7
grasswire May 2013 #27
Bjorn Against May 2013 #8
uppityperson May 2013 #9
rufus dog May 2013 #11
yodermon May 2013 #12
Bjorn Against May 2013 #13
yawnmaster May 2013 #14
Bjorn Against May 2013 #15
yawnmaster May 2013 #16
Bjorn Against May 2013 #18
yawnmaster May 2013 #20
Bjorn Against May 2013 #24
magellan May 2013 #17
Bjorn Against May 2013 #19
spanone May 2013 #21
Bonobo May 2013 #22
Bjorn Against May 2013 #25
Dragonfli May 2013 #23
MannyGoldstein May 2013 #26
grasswire May 2013 #28
Bjorn Against May 2013 #29
burnodo May 2013 #30
Bjorn Against May 2013 #32
cali May 2013 #31
Bjorn Against May 2013 #33
cali May 2013 #34
Purplehazed May 2013 #35
Bjorn Against May 2013 #40
One_Life_To_Give May 2013 #36
John1956PA May 2013 #37
Bjorn Against May 2013 #39
LineNew Reply Bayes Rule is useful here.
BlueCheese May 2013 #38
Please login to view edit histories.