One thing to keep in mind is the type of subscribers AT&T is bringing in... They're likely losing non-data (high end) subscribers and they are certainly bringing in high revenue subscribers (since iPhone requires a 3G data plan).
This whole post is flawed - a mystery wholesaler from T-Mobile giving accurate stats on ATT/Apple flag ship product, yeah right. - based on only 1,000 ports? - info of such a small base compared to the exact small base last year?
For all we know all those incoming T-Mobile accounts were people using unlocked 1st gen iPhones sold by people upgrading to 3G/
You can't nitpick about "only 1000 ports" being used to draw a conclusion- pretty much every survey samples around 1000 people because that gets good accuracy and still allows for easy-to-use distributions to describe the data. A bigger problem (which you mention first) is that we're only taking data from one T-Mobile distributor.
The biggest problem is that you only tell me that the percentage of people porting from T-Mobile to AT&T is lower than it was last time. How did it change in absolute terms? Had there been a trend of people porting less to AT&T now than before? I also would imagine that people porting to T-Mobile care much less about data and, therefore, probably have lower monthly payments than those who ported away from T-Mobile. None of this is answered by the one random wholesaler.
@mac404: You're mistaken about being able to draw a conclusion from this sample. Yes, a sample size of 1,000 is plenty to draw a statistically meaningful conclusion, but what you're missing is that it MUST be a truly RANDOM sample of the entire universe that the conclusion is being drawn about. This is not a random sample. We have no way of telling whether it's representative of the wider picture at T-mobile or not. So it's insane to try to draw conclusions this way, and it's ignorant of research methods to pretend that any 1,000 results (of unknown origin) can accurately say anything about the bigger picture.
Now that we've thrown 'em off the trail, use the form below to get in touch with the people at Engadget. Please fill in all of the required fields because they're required.
One thing to keep in mind is the type of subscribers AT&T is bringing in... They're likely losing non-data (high end) subscribers and they are certainly bringing in high revenue subscribers (since iPhone requires a 3G data plan).
This whole post is flawed
- a mystery wholesaler from T-Mobile giving accurate stats on ATT/Apple flag ship product, yeah right.
- based on only 1,000 ports?
- info of such a small base compared to the exact small base last year?
For all we know all those incoming T-Mobile accounts were people using unlocked 1st gen iPhones sold by people upgrading to 3G/
@andrew - word.
Let's extrapolate a tiny amount of data to an enormous conclusion!
You can't nitpick about "only 1000 ports" being used to draw a conclusion- pretty much every survey samples around 1000 people because that gets good accuracy and still allows for easy-to-use distributions to describe the data. A bigger problem (which you mention first) is that we're only taking data from one T-Mobile distributor.
The biggest problem is that you only tell me that the percentage of people porting from T-Mobile to AT&T is lower than it was last time. How did it change in absolute terms? Had there been a trend of people porting less to AT&T now than before? I also would imagine that people porting to T-Mobile care much less about data and, therefore, probably have lower monthly payments than those who ported away from T-Mobile. None of this is answered by the one random wholesaler.
1000 samples is good if you're using a simple random sample. I want to say this is a sample of n = 1 store, not random at that.
@mac404: You're mistaken about being able to draw a conclusion from this sample. Yes, a sample size of 1,000 is plenty to draw a statistically meaningful conclusion, but what you're missing is that it MUST be a truly RANDOM sample of the entire universe that the conclusion is being drawn about. This is not a random sample. We have no way of telling whether it's representative of the wider picture at T-mobile or not. So it's insane to try to draw conclusions this way, and it's ignorant of research methods to pretend that any 1,000 results (of unknown origin) can accurately say anything about the bigger picture.