Last night ended with a conversation much like ones we’ve been having for the past two months: the broad topic was the iPhone 3G, and the specific topic was a report that 91% of Japanese consumers were not interested in buying the device. After three minutes of discussion, we decided not to run this report—a decision we’ve made quite a few times over the past two months—and with that, we called it a night. The consequence was that our news column wasn’t as packed with fresh content, but we felt that we’d made the right choice.
You might be asking yourself why we wouldn’t want to run just about any story with the words “iPhone 3G” in the title. After all, it’s obvious that our readers are interested in the device—64% of the people responding to a recent iLounge poll said that they were planning to buy one—and of course, more iPhone 3G stories equal more page views, right?
Our reasoning was very simple: the story, like many others we’ve passed on, was ridiculous. Amongst other red flags, the survey was taken days before the iPhone 3G was even announced. It was based upon only 402 participants. And it was limited to a specific demographic (20-49 year old Internet users).
In other words, when 9% of respondents said that they planned to purchase something, they were being asked about a device they didn’t know about, and a total of 36 people said yes.
What was fascinating to us was the way that this survey story was processed and regurgitated elsewhere. Some repeated the original story’s negative conclusion verbatim: “91% of Japanese will not buy the iPhone.” Others tried to spin the findings in a positive way, claiming that the survey meant that 9% of Japanese will buy the iPhone. Given Japan’s population of 127 million people and 103 million mobile phone subscribers, the survey supposedly revealed to certain readers that 9.3 million of the country’s mobile phone subscribers were planning to buy iPhones. All this, based upon 36 people saying yes.
While surveys are unquestionably useful, they require fairly large sample sizes in order to be accurate. To poll the entire Japanese mobile phone user base with 99% confidence, with a desired accuracy of plus or minus 3%, you’d need to sample around 1,850 people. Increasing the accuracy to plus or minus 2%, you’d need to sample 4,160 people.
The Japanese iPhone survey didn’t even come close to those numbers, nor did it attempt to broadly sample the country’s population. Something’s fishy when an iLounge survey has a sample size of over 6,000 people and a purported survey of the entire country of Japan has only 402.
There is, of course, an standard excuse of sorts for publishing a “story” like this: “Who really knows whether it’s right or wrong? It’s just one more piece of possible information that might potentially lead to some greater understanding.” In a word, that’s irresponsible. The old “throw everything against the wall and see what sticks” trick only increases the amount of junk that’s out there—we’d dare say that it’s about as contrary to Apple’s approach to doing things as is possible.
For that reason, we’ll skip the lengthy explanation of why we didn’t post all of the phony iPhone 3G images that appeared across the Internet in the runup to the product’s debut. The short, simple answer is that we knew they were fakes and didn’t want to waste your time with them. The only ones we ran, along with explanatory text, were the real things. We knew it would have been fun to discuss all the things that people hoped would make it into the device: that’s why we held a concept contest to let readers express their creativity.