A stupid survey

Filed under: news @ 13:56

I’ve seen it floating aroung the internet here and there over the past several days, so I finallly decided to look at it: Intel’s ranking of the top 100 “Most Unwired Cities Survey“.

‘Unwired’, here, is taken to be a good thing. The survey attempts to judge the wirelessness of those cities it ranks: wireless hotspots and the like. The problem is that it doesn’t do what it says it does.

Firstly, it doesn’t rank cities as the title of the survey suggests, but metropolitan areas instead. This isn’t that big of a deal, and, indeed, the explanatory notes make this fact explicit, but it’s still sloppy. More importantly though, this does not rank the top most unwired cities.

The survey describes its task three separate times:

  1. In its title: “Most Unwired Cities Survey”. This is false.
  2. “Intel’s 3rd Annual ‘Most Unwired Cities’ survey ranks the top 100 U.S. cities and regions for the greatest wireless Internet accessibility”. Sloppily written but one could argue that it’s true.
  3. “Below are the top 100 cities and regions in the United States for wireless Internet accessibility, where you can unwind and unwire:”. Also false.

Given the above quotes and their respective veracities, can you guess where exactly Intel tries to pull the wool over people’s eyes? I’ll give you a hint, they attempt to trick you with marketing, not statistics; it’s in the language they use, not the numbers they produce.

Answer: They only surveyed the largest 100 metropolitan areas in the U.S. and there is no reason to expect that the set of the largest 100 metropolitan areas is the same set as that of the top 100 most unwired.

Allow me, if you will (and what choice do you have), to illuminate the problem with an example. If someone gives you five apples (and you know nothing about the original selection process this person used to cull five apples from the millions that exist) and tells you to arrange them in order from biggest to smallest, you would never think to say “I’ve just ranked the five largest apples”. Or, rather, if you would think to say that, there’s probably a career ahead of you with Intel’s news department. Intel did not rank the “top 100 most unwired cities”, they ranked the largest 100 metropolitan areas by their unwiredness. Big difference.

Cruising the internet, there are a lot of people who recognize this difference, but there are a vast number, still, that got duped.


    gv 06.22.2005 @ 17:04

    Below I have ranked the top three websites*:
    1. apparenthorizons.com
    2. langabi.name
    3. blogwaffe.com

    *Rankings were decided from an extensive survey of three websites including those listed above.

    MDA 06.22.2005 @ 17:07

    I’m in the top three! Booyah!

    L'el 06.22.2005 @ 20:02

    Well I don’t feel too bad considering I have a newer blog; I’m sure I’ll crack the list next time. In the meantime, how hip is it that I discovered and was a regular reader of ALL THREE top three websites before the official ranking came out? Damn, what good taste I have!

    Of course, now that this knowledge is “mainstream” it’s time for me to stop reading these things and move on to the next big thing: which must be returning air conditioners to Best Buy. A new survey coming out, well, right in this comment section, says that returning air conditioners to Best Buy is the hottest past-time for twentysomethings on Tuesday afternoons. Wheeee.

    Adam 06.23.2005 @ 00:43

    This made me think of a T-shirt that the have over at Thinkgeek. It’s kind of morbid, but rather accurate. “PEBKAC”, for those who don’t know, stands for “Problem Exists Between Keyboard And Chair”. In this case, it means that the problem isn’t Intel’s misleading survey, but the idiots that buy it. As long as there are idiots in the world, there will be marketing-based surveys rather than fact-based information.

    I also have a question. Why are you getting so worked up about this Mike? It’s not as if this is the first time something like this has been put out there. Nor does it have a huge impact on you. Were you planning on moving to one of these places? I honestly doubt that a survey like this will have any measurable impact on the housing market of any of these areas.

    Oh, and your preview button no longer works in Opera.

    MDA 06.23.2005 @ 01:15

    I could not disagree more. The problem is Intel’s. If you were to publish a scientific paper which was sloppily written at best and deliberately misleading at worst and were subsequently called on it, you’d never get away with saying “oh – the fact that you didn’t correctly interpret my results is your fault. You should have more carefully analyzed my numbers and methods rather than believing my description.” That’s just absurd.

    (Even better: “It’s the public’s fault there we’re stuck in Iraq under false pretenses. If the people of this country had only critically examined these intelligence reports and considered the means by which the data were obtained, they’d have known there were no WMDs, and we wouldn’t have gone to war in the first place. Rather, WMDs would not have been our excuse.”)

    As for why I’m worked up about it, I’m not really. I happened to see someone misrepresenting the survey and that made me realize just how bad Intel’s presentation of it was; this person wasn’t stupid, but overly trusting. I just thought I’d point the issue out, even if it doesn’t impact me (or even anyone reading this) personally.

    That said, it’d be perfectly reasonable to get worked up. People should read more critically even if it is Intel’s fault there were errors. More importantly, Intel should recognize that they’re publishing bad science. Bad science being propagated by a well known, well respected and powerful company is always reason enough to get vocal.

    Plus it was a slow day.

    (The Opera thing is bad news. It’s not something I’m going to have time to look into anytime soon.)


    […] A friend of mine saw the article and decided to challenge it. In mdawaffe’s article, “A Stupid Survey”, he says: ‘Unwired’, here, is taken to be a good thing. The survey attempts to judge the wirelessness of those cities it ranks: wireless hotspots and the like. The problem is that it doesn’t do what it says it does. […]

    jim 09.12.2005 @ 04:34

    do u expect me to read all of that forget bout it

    Anonymous 11.26.2005 @ 19:52


    jimmy 12.04.2005 @ 06:33

    you are the most retarded website i have ever seen!

    jjk 12.04.2005 @ 16:33

    Those last two comments are priceless.

© mdawaffe (Michael D Adams) - Powered by WordPress - Full Credits