It's not a "sound" study for what gsaldivar is trying to use the data for. You are arguing something not just on a tangent but completely different from what gsaldivar and I have been arguing about. Gsaldivar is trying to say the data contained in the Squaretrade report is indicative of the iPhone market when it's not. By it's very nature everything in the Squaretrade report can be true but not applicable to the iPhone 4's as a whole.
The Squaretrade report is in essence anecdotal evidence. It is no different from me telling you that out of all my friends and families with iPhones I've only seen two with broken glass. It's roughly 25 people that I know with iPhones. Does that mean 8% of the people with iPhones have broken glass? Of course not. That's a ridiculous conclusion to make. The data is true but at the same time it's not applicable to the iPhones as a whole. You could pick another guy who knows 25 people with iPhones and he might know 7 of them with broken glass. Does that mean 28% of iPhone users have broken glass issues?
Squaretrade may have a higher number of iPhone users to draw from but it's the same principal. Another warranty company could just as likely track 20,000 iPhone users and have only 2% broken glass issues. While Squaretrade may have close to 5% of people reporting problems with iPhone 4 broken glass, it might only be 2% of the total number of iPhone 4's having broken glass for whatever various reasons.
And the 80+% more issues with broken glass issues from 3GS to 4 is a number that has almost zero relevancy when you look at the absolute percentages of 3GS's and 4's with broken glass. Again, it's like how the iPhone is "losing" market share but nearly doubling iPhone sales from the 3GS to the iPhone 4.
Actually I agree with you saying the iPhone 4's are more susceptible to glass breakage. However, you're arguing that the Squaretrade report (I can't really call it a study) is indicative of iPhone 4's as a whole and I'm saying we can't come to that conclusion.
So I do think the iPhone 4's are more susceptible to glass breakage simply because it uses larger pieces of glass and two per device vs one. I also can't agree with your conclusion that the iPhone 4's glass breaking is problematic enough to warrant class action status. I'm saying there is not enough sound data to say one way or the other.
I mean, let's be honest here, glass breaks even if it's better quality glass that is less susceptible to breakage. That's just the nature of the material. If you drop that it just right, it's going to break.
Apple has never claimed the iPhone 4's glass is unbreakable. Nor has it been proven that there is an exceptionally high rate of iPhone 4's with broken glass that indicates a hardware/design issue. So unless someone out there can prove there is some sort of defect that is causing the iPhone 4's (this Squaretrade report isn't going to cut it in court) to have a higher than expected number of iPhones with broken glass, then this is a frivolous lawsuit. It's a money grab pure and simple.
I don't want to nitpick details too much. Bottom line for me, this "study" isn't conclusive. It's anecdotal in nature. The sampling size is quite large but it isn't enough evidence to show there is a hardware problem causing glass breakage. Nowhere will you find Apple saying the glass used in the iPhones are unbreakable.
Unless someone has some new evidence such as an actual study on the percentage of iPhone 3GS's and iPhone 4's with glass issues that is more sound in nature or a report that proves there is a report saying there is a hardware defect that is causing higher than normal glass breakage on iPhone 4's then I'm sticking with my opinions. If you still agree with what I'm saying, we'll just have to agree to disagree.
Lies, damned lies, and statistics.
Do you even understand what "Anecdotal evidence" means or do you just like saying it to sound smart? Have you even had a single class or used actual statistics in the workplace?
The expression anecdotal evidence has two distinct meanings.
(1) Evidence in the form of an anecdote or hearsay is called anecdotal if there is doubt about its veracity; the evidence itself is considered untrustworthy.
(2) Evidence, which may itself be true and verifiable, used to deduce a conclusion which does not follow from it, usually by generalizing from an insufficient amount of evidence. For example "my grandfather smoked like a chimney and died healthy in a car crash at the age of 99" does not disprove the proposition that "smoking markedly increases the probability of cancer and heart disease at a relatively early age". In this case, the evidence may itself be true, but does not warrant the conclusion.
So what is an "anecdote"?
"An anecdote is a short and amusing or interesting story about a real incident or person. It may be as brief as the setting and provocation of a bon mot. An anecdote is always presented as based on a real incident["
An anecdote is a single, or even a small group of people, saying "ohh yeah, I dropped my phone and it broke". It's a story, not a statistical study.
So you're telling me that a sample size of 20,000 equates to 20,000 anecdotes?
Are you fucking serious?
A sample size of 20,000 is more than enough to extrapolate a statistically significant result from a population. In fact, considering the size of the iPhone4 population, I would say that you could probably derive significant confidence in the result of such a statistical study.
The higher number of samples the more confidence you can have. Sure, different warranty companies might report different numbers, however, the law of large numbers says that as the sample size grows the numbers should revert to the expected mean. Thus, if Squaretrade WERE outside of the "mean" with it's 20,000 samples, as soon as it hit 40,000, you would see the difference. Even at 20,000, the chances that another warranty company would experience different results is slim, at least with any large margin of errors.
What you're effectively doing is taking a STATISTICAL STUDY and re-labeling it as a non-statistical study merely because YOU don't think it is statistical and want to brand it as "anecdotal", when, in fact, almost any person with a statistical background would agree that this is a decent representational sample that is likely not biased by selection issues or regional issues.
Similar sample sizes for far larger overall populations can produce, more or less, the same results. For example, most political studies are of only 20-30k people, if not less, but are very indicative of the US population as a whole, especially once you consider the voting results. Now 30k people is less than .01% of the US population, but if they are drawn from many states and called randomly, then that is not 30,000 "anecdotes", it is a statistical study with a confidence interval of maybe 90-95%.
But hey, don't let fact stand in your way.
