Wednesday, June 28, 2006

Thumbs down for the dishonest websites amongst the Mac Web.

Generally I like TUAW (The Unofficial Apple Weblog), but a survey revealed yesterday was an example of very poor reporting.

The results of our MacBook discoloration poll are in; 20% of TUAW reader's MacBooks exhibit discoloration. It's hard to judge how accurate these results are, but if the real numbers are anywhere near these levels then Apple has a serious problem on its hands.

Fortunately, there are signs that Apple has taken notice: the creator of recently managed to get Apple to agree to replace the plastic on the top of his MacBook's casing. That's a big step, but I'm sure many MacBook owners won't fancy calling up Applecare every two weeks after their replacement starts showing stains again. Clearly, a more permanent solution is required.

A few comments have questioned the accuracy of this poll within the TUAW forums:

You write that "20% of TUAW reader's MacBooks exhibit discoloration" -- in reality, you can only say that 20% of TUAW reader's who responded to this poll indicated that they have discolored Macbooks -- a very different thing, with far less certainty. Jumi

Another commenter posted this irrational conclusion in reponse:

Jumi, frankly 1,500 reponses is a large enough sample to constitute representative poll, which means that this figure wouldn't change much if I applied it to all our readers.

From Wikipedia: "Opinion polls are usually designed to represent the opinions of a population by asking a small number of people a series of questions and then extrapolating the answers to the larger group." Conrad

First of all ... an online poll can never be scientifically accurate. Apple's competition, Apple Haters, and punk teenagers will surely skew the results. TUAW is big enough to attract attention to a poll like this and thereby having the results skewed by one of the aforementioned.

Second, online polls are different from a random walk up survey. Pollsters with a problem most likely sought out "discoloration of their MacBooks" through a search engine or it caught their eye in a headline. A random walk up survey of MacBook owners would certainly prove the skewing of such a poll as this.

Third, TUAW is not big enough to be respected. If Apple Computer were conducting this poll, less people would be inclined to lie. You can also be assured that Apple probably has a few noted IP addresses and email addresses that they could see if someone were manipulating the results for one reason or another. ONLY a poll that got valid email addresses from the respondents could be accurate.

And by the way, I've heard that high end car finish/scratch remover not only removes the discoloration but prevents it from happening again.

No comments: