The
Loveline
Companion

Home > Forum > Archive > January 2007 > Semi scientific loveline metanalyisis

Login

Semi scientific loveline metanalyisis

  

yevishere

+

Tuesday, January 2, 2007 at 6:53 PM

Hi All,
I am a statistics guy and I've been listening to loveline a lot lately, because I was able to download a bunch of shows. Anyways, I've been thinking, with all these recorded episodes I have, I could build a database, something like this
Show_date Age_of_caller Sex Name Problem Age_abused Times_abused type_of abuse
01-01-07 15 F candy masturbation 10 2 sexual


Anyways from this you can do a regression analysis to figure out which factor has contributed to the problem if any, for example, sexual abuse can be associated with masturbation, ect with probability X. The biggest problem is data entry of 500 or more points(callers), to make this viable. If it works it might prove a lot of things that Drew and Adam have been saying over the years.

What are your thoughts, have anyone heard of this idea before, is there a database out there I could use? Is it a worthwhile project? If its a good project, would anyone want to help with accumulating data points?
Thanks,
Eugene

yevishere

  

anobody

+

Tuesday, January 2, 2007 at 10:59 PM

It's a cool idea but there's one huge show stopper - the calls aren't a random sampling of the population - they're selected for whatever the producers, screeners, and hosts thought would be interesting, and self-selected based not only on who had a specific issue, but also who thought calling Loveline would be a good way to address their issue.

In order to correct for all that skew, you'd need to know so much about the population, you'd already have the results.

anobody

  

bguirk

+

Tuesday, January 2, 2007 at 11:02 PM

Yeah--Adam fucks up the sample when he yells at Bryan to bring him more big boob calls.

bguirk

  

anobody

+

Tuesday, January 2, 2007 at 11:14 PM

You kid - but you're absolutely right (although you could correct for that bias pretty easily).

I was thinking more along the line of "oh we've taken too many molestation calls tonight - lets get something different", or "we've never had one of these bizarre threesomes with an emu and a bull whip - lets take that".

anobody

  

ZT-In-Thought

+

Wednesday, January 3, 2007 at 12:20 AM

LET'S DO IT!!!

ZT-In-Thought

  

unpwnful

+

Wednesday, January 3, 2007 at 2:47 PM

http://www.lovelinearchive.com/

for the downloads, you could divide up the shows by date and have people sign up here and then report back when they have collected the info.

unpwnful

Home > Forum > Archive > January 2007 > Semi scientific loveline metanalyisis ~ Post a Follow-Up