It's so true -- you tend to end up with the opinions of the two most dominant people because the others just back down.In the late 1990s-early 2000s, I facilitated the online focus groups for Knowledge Networks. Most of their polling was by survey of their panels, and they had placed WebTV Plus boxes with each of them for that purpose (put everyone on the same platform, and one that required no computer or technical skills). We were consulting for them on other things, and at the time we ran a weekly chat+radio show of our own. It was something like Blogtalk Radio, except that we used IRC for the chat and broadcast in RealAudio, which WebTVs could also use. We made some custom interfaces so that people could easily chat and listen.All of this may sound ordinary, but we started doing it in 1997. As far as I know, we invented the format, as we certainly weren't aware of anyone else doing anything like it. So that's how we did the KN Focus Groups, using the same set-up (and a custom interface), and they were really effective. We got a very clean chat transcript of question and answers, since all of the coaching was done in audio (I would paste the question into the transcript in a different color so that they were distinguished). We also had the ability to shut people up without their realizing that they had been, private message a participant, and wrote a nickserver that would cloak the actual chat name with a pseudonym for privacy (because WebTV normally used your email address as your chat nick). Best of all, though, was that people couldn't be drowned out or intimidated by someone who was more opinionated. The results were quite good.And sometimes, quite fun. My very favorite focus group of all was the set we did for Smirnoff high-proof vodka, which became something like a party that nobody wanted to end.But normally, focus groups tell you what you want to hear. That's not necessarily what you NEED to hear.
Focus groups are quite likely to lead you astray. The social dynamics produce weird results, as shown by[/break1]youtube.com/watch?v=624FxhJlVM0]https://www.youtube.com/watch?v=624FxhJlVM0
Crap, that rules me out.Hell, intelligent, well-educated available men are hard to find in general, lol. [highlight]If one adds in the requirement that they not be assholes[/highlight], the world becomes a small place.
Interracial dating/marriage aside, this clueless idiot does not seem to recognize the racism in his statement that intelligent, well-educated black men are hard to find.
No doubt the advertising agency checks the latest census figures before selecting which ad airs where. After all, it is the Department of Commerce that is in charge of the census. Most of the questions are for the purpose of commercial interests and they sell the block information..without the names and SS# of course..but all the demographic and economic information....block by block. Have you ever looked at your block? How hard is it to figure out who is who? #-o Just something else to data mine. I just let them guess at mine. No forms from my address except what they make up. It's just about all wrong information. Only people like Sankey, Hendershot and other PI's hired by O'rly would pay too much attention to those data bases. GIGO. Credit reports. Same thing. Tax Assessor rolls. Same thing. They are all full of errors. Not lies, ...sometimes lies....mostly errors. Accurate data are really hard to come up with unless you verify, cross-check, triangulate, and reconfirm from many angles.
Don't know what other areas the State Farm ad is running in but it's on constantly here.
Source: Bureau of Statistical Analysis Management
-George Washington, 1801.
So you might like to think.* I might agree with you if I had not accessed the information for my little block and was able to figure out pretty much who was who in which household. If I had used other public resources such as tax assessor records, white pages, court filings, internet postings, I could have pin-pointed everyone with close to 100 % accuracy. Granted I have not attempted this recently with the latest census, but I did with the last one. * using the stern tone of the character in the BBC version spoken to the camera explaining the inner machinations of governmentAs for sampling, I remember those discussions. Some groups were for and others against sampling for their own particular interests. If a sample is too small, it is not worth much. The sample could be big enough to include almost everyone and still be called ''a sample.'" I have this same discussion about the meaning of 'random'' a few days ago. You could choose random applications to check for veracity of information supplied on the application. The theory is this deters people from trying to cheat and entering false information. Of course, this only works if they know people get caught. Random could mean anything more than zero and less than 100% of the applications would be verified. It's like the old Skinner experiments with chickens pecking a button for food that was later applied to slot machines. If the reward is random but too infrequent, the chicken gives up after a while.
The Census has always restricted information released at the block level and has always suppressed information if the cell size is too small. The old samples were never constructed to be representative of blocks, so most of the sensitive data were never available at the block level. Today's Census does not even collect the detailed economic, occupational, and educational data. That is done by the American Community Survey, which is a sample, not a census. Data cannot be generated at the block level from the ACS.There are all sorts of private firms that pretend to have such data.
Source: Bureau of Statistical Analysis Management
-George Washington, 1801.
This risk of re-identification has risen substantially with the availability of external databases, some public-access and some not, some useful and many not. It was probably always possible for a really well informed neighborhood snitch to make good guesses about who was who in the Census data, but now the problem is serious enough that we may lose access to some of the small-geography aggregate Census data to which we once had access.
So you might like to think.* I might agree with you if I had not accessed the information for my little block and was able to figure out pretty much who was who in which household. If I had used other public resources such as tax assessor records, white pages, court filings, internet postings, I could have pin-pointed everyone with close to 100 % accuracy. Granted I have not attempted this recently with the latest census, but I did with the last one.
This risk is discussed in a long series of reports of the Committee on National Statistics of the National Research Council, published by the National Academies Press. The aim of those reports has been to provide advice to the government on how it can provide access for the public to the data for which they have paid while still abiding by the terms of Title 13, which protects the privacy and confidentiality of the data. Expanding Access to Research Data: Reconciling Risks and Opportunities (2005) provides a good overview of what was then known and could then be anticipated.
We may end up in a world in which none of the data released are real. They will instead be synthesized or modeled data in which the statistical results would be "the same" as if you were working with real data, but there are no real people, families, households, or dwellings in the data set. I have opposed that but see it coming.
Bah! Oats are good for you!
I saw the commercial again today while working out with my personal trainer, Vinnie. The only problem we found with the commercial was that it encouraged eating grains. That is a problem in 2013.
Who is online
Users browsing this forum: No registered users and 1 guest