I’ve been looking at doing some book-marking for a few of my sites but wasn’t certain if it would yield any
positive results. Appreciate it.

Either way, it is only fair to let you know that I have stereotyped you as an intentional-time-waster and as a result I will probably not respond to your parting ad hominem attack. I’m not inclined to have my time wasted. Since you were not inclined to engage in the ongoing conversation anyway, hopefully you will not have invested so much in the ad hominem attacks that this comes as an emotional blow to you. May your parting nag provide whatever emotional healing you may need and sustain you in your search for the next conversation or person unworthy of you.

I’m not an Apple stock owner, but that’s mostly for lack of cash; and if I was an Apple stock owner I would offer full disclosure when I posted comments online … I think it is the respectful and ethical thing to do in a public discussion.

Which argument was that? You don’t specify, and I made several points, at least three major ones.

Also, what prejudice are you referring to? I do have a prejudice against sloppy use (or misuse) of the English language and the manipulative use of statistics. Not sure there are any other “prejudices” relevant to the now very old discussion above.

Been surprised at the lack of discussion about two other ways Steve might have used statistics to mislead.
—————————-
1. He only reported AT&T return “rates”. Why not report Apple Store return “rates”. This is a glaring omission and leads to only one logical conclusion: Apple has had more returned than AT&T has, perhaps a lot more. Also given that we are talking about a rate, we need to know the unit time. Wouldn’t be surprised if a different unit of time was used for the 4 vs. 3GS. If true, the direct comparison of these two “rates” would be a very intentional lie. Like saying I’m 175 tall and you are only 5.75. Basically useless information to an educated person, but clearly implying that I’m taller to the uneducated. I may be taller but you can’t tell from these numbers. As soon as I clarify that I’m reporting my height in cm and yours in feet, an educated person can do the math and figure out that we are actually the same height. There is no such thing as a rate without a unit time, though we hear the term used without specifying a unit time constantly in the media and Steve very likely took advantage of that fact. In spite of the true popularity of the 4, I’m inclined to believe Steve told a lie on this one. He didn’t explicitly say anything false. But he very intentionally misled.
———————————————
2. Likewise, when he described the .55% Apple Care complaint “rate” as low, he did not specify the time window. He just said “first few days”. A rate is a unit of measurement that includes (by definition) a unit of time. He did not give us a unit of time, so the .55% figure is utterly meaningless. He could be talking about the first 2 days of sales for the iPhone 4 and saying these figures are much smaller than what we typically see after selling a device for a month. Well no kidding! Again, since he didn’t specify what time units he was using in his comparison, he did not explicitly give out false data. But by comparing two values on different scales without clarifying that they were not measured on the same scale, he very purposely misled us. Again, I don’t know for sure these were different scales, because he never told us what the scales were. But even the less cynical half of me says it is very likely that he mislead us all with these unitless “data”.

Why am I so convinced that he misled us? Because he was not willing to explain the real difference between the iPhone antenna issue and other antenna issues. By this point, he has to know what these differences are. With 18 PhDs and all, its just not possible that he doesn’t get it. He wouldn’t have offered cases as a fix if he didn’t get it. But he still won’t come clean, and Apple continues to post videos implying that the problem is the same for the 4 and every other phone on the planet. That’s simply not true and his engineers know it, and he knows it. Now it may be true that the 4 has better reception than other phones in many scenarios. But this is something he could have explained. Instead they have shown that they are willing to lie to get through this. Would I do the same in his place? Possibly if I believed the success of my company and my life’s work depended on it. Otherwise, no, I’d just come clean, explain the true pluses and minuses of the design, and build that trust he claims he has earned but actively demonstrates he has not. The fact that they won’t come clean makes it seem more and more likely that the design was not a proactive choice but a genuine mistake on the part of incompetent engineers. I find that hard to believe really. But the use of misleading info to patch things over seems to suggest the truth would be worse than a string of lies.

Would you say iPad merits the 8th statue?

Or is the jury still out?

And what we care about is not so much what happened as what is likely to happen going forward.

Another context is WHY the call dropped. If 2% of calls drop overall, and half are due to AT&T congestion while the other 1% are due to losing a weak signal, a 1% increase would suggest that the iPhone4 is twice as likely to drop a call due to weak signals. If 98% of calls originate in strong signal areas, and 2% in weak signal areas, that’d mean the weak-signal-failure-rate went from 50% to 100%.

My personal experience suggests it’s not so dramatic. I think my iPhone4 has better weak-area performance than my 3G. Maybe it’s because I find it uncomfortable to talk when my fist is wrapped around the phone: my knuckles press against my cheek; the square edges make me favor holding it by my fingertips. So many variables that you’d like not to have to wonder about, even if the statistical musings are fun for a bit.

Appreciated your post, but had to comment on an error that several others have made when discussing Jobs’ data:

The increased number of dropped calls from 3GS to 4 is absolutely statistically significant. It would be significant no matter how small or large the number because AT&T is not sampling a subset of calls: They know the exact number! This is an actual count made from every single call in the country, not a sampling. If it was a sample, statistical tests could tell us with a precise degree of certainty (say 95% certainty) whether the difference in the sample was also a difference in the population at large. We don’t have to make that inferential leap from sample to population, since we have the population data here. We can say with 100% certainty this difference is statistically significant (or if you prefer: “real”).

The more important question is whether number of dropped calls is large enough to be meaningful. There are two problems in answering this question. One is that statistics cannot tell us what number or effect size is meaningful in the real world. This depends on one’s perspective (or circumstances) I’m afraid. If you need that “one” dropped call in an emergency situation, then it is very meaningful. If you can deal with waiting and then redialing once or twice a day, then even a very large “effect size” might not be all that important. The second problem is that Jobs did not give us the actual effect sizes. If 1/100 calls being dropped amounts to too large of an effect size for you, then one would assume the 3GS exceeds this threshold too … but we have no way of actually knowing. If your threshold is closer to 1/10 calls, then we simply do not know if any iPhone will work for you. They might both work. Also possible only the 3GS would come under your threshold. But Apple and AT&T didn’t give us the data to make that decision (as Steve said: he “totally agreed” that they shouldn’t).

Long story short, don’t use the language of statistical significance to make this seem like it is an insignificant number. It might be. It might not be. But it has nothing to do with statistics (descriptive, inferential, or otherwise).

Video calling got off to a bad start in Europe despite it being one of the major advertised features of 3G because of several issues. Video calls between different operators were not possible initially. Then price gouging made the video calls far too expensive. But now these are not an issue and many plans now include free video call minutes just like voice minutes. But even when calls are effectively free they still account for just a percent or two relative to voice calls. From what I have read elsewhere the figures are so bad that most network operators have stopped separating out statistics for video calls.

The only advantage I can see that FaceTime might have is call quality. Because of the bit rates involved in 3G video the quality isn’t so great. But I don’t think that accounts for the lack of use. I just think that in general people don’t care for video calling outside of a few special use cases.

Still, it is amusing to see Apple’s FaceTime advertisements and realize that the European network operators were using basically the same themes in their 3G video calling advertisements in 2004.

Sure he could have. Giving out limited edition iCondom 4′s would have shunted this story off the main line. However, I kind of like the fact that Jobs didn’t get it right. It makes him more human. We know the stories of his perfectionist behavior and anyone suffering this probably doesn’t see it as an endearing trait. The fact that the product is more often than not a winner attests to the value in this. If Jobs were also a “common man” enough to see the easy way out of this, it would also mean he didn’t have the pride and hard-headed attitude to make sure his products went out the way HE wanted them.

So how about move on to something more important and interesting? I’m really tired of the fanboys…how about go explore RiM’s new OS or something useful?