[quote name=\'Dan Sadro\' date=\'Aug 2 2003, 08:33 PM\']It's not a flawed system at all... [/quote]
We just read why it's flawed. Only a certain kind of person will participate. Sure, the sample selection is great methodology, but many of those selected never participate.
I worked for Simmons Market Research which rates magazine readership. They use a one-time personal inverview involving recall of mags and specific ads. We also gathered a lot of demographic and psychographic info to help answer \"nationally, who reads each of these mags\". Interesting stuff. Similar to Nielsen's methodology, 20 households were randomly selected from a cluster (smaller than a zip code), and the local field interviewer tries to get 15 or more of those households to respond to make that cluster usable. The interviewer gets bonuses for each household above the 15 that he can get. I was so adept at getting into tough homes that they sent me around the country to fill in households from under represented clusters.
I can tell you what kind of person lets you into their home easily, who requires some very creative terms to overcome their objections, and who resists tooth and nail, despite the gifts and cash to entice their one visit involvement. Compare that to the much more imposing requirements to be part of Nielsen's most prized sample - the people-meter homes providing the national \"overnights\".
Multiple phone and in-person interviews, installers to bring in the hardware and connect the rig to your phone line, entering demographic info of everyone in the household into diarys, promises to press your individual button regularly so that they know which member(s) of the household are watching, more phone contacts, allowing an installer back in your home to remove the rig, etc.
The point? Only certain kinds of people are willing to grant you a single in-person interview no matter how much you play Monty Hall with gifts to keep their interest. Now add all the crap required to be part of that Nielsen sample and you can immediately count out busy, upperclass professionals - no doctors, lawyers, senior management types, etc.. Count out anyone living in an expensive home - they don't want strangers traipsing through their private domain unless you're talking thousands of dollars. Forget homes in the central city - they are sure you're trying to rip them off in someway, from a high pressure sales pitch to casing their homes for a later burglary. Unless it's a dirt poor household in an urban area - they'll say anything they think you want to hear for $10 cash and trinkets for their kids. Forget any of the surprisingly high number of households where an illegal lives, or where someone in the home has an outstanding arrest warrant or is a parole violator. Same for many of the owners of homes who have converted their garage to an apartment without benefit of proper zoning and/or building permits. There's also no way with potential participants of a specific psychographic profile - the paranoid and government fearing, as well as the just plain rebellious or aging hippie. Getting the picture?
My favorite was a guy in his 30s who was living with 2 underage girls. He was providing them with drugs, screwing their brains out, and pimping them. Ya gotta love it! Yea, I got his interview and household demo info, including the girls - I didn't need their names, only his. I ended up staying for dinner and well into the evening.
Still reading? Get a life! ;-)
While, as fans, we'd LOVE to be part of a Nielsen survey, for most random folks this is as interesting as a survey about root canal, and they just don't care nor want to submit to the time and trouble. So if only certain people can be coerced to participate at all, and many of them fill out the entire diary when the letter comes to send it back, I submit that this is not even close to a sample of the population. Flawed? Hell yes.
It's like navigating Vermont with a map of New Hampshire. It's the only map you have, so you use it as best you can.
Randy
tvrandywest.com
