American Consumer Satisfaction Index (ACSI) Report Evaluates ConsumerSatisfaction with Wireless Carriers. T-Mobile and Verizon top ratings, Cingularand Sprint tied at the bottom.

[POSTED TO alt.cellular.cingular - REPLY ON USENET PLEASE]

In <[email protected]> on Thu, 18 May 2006 15:33:24
-0700, SMS <[email protected]> wrote:

>[email protected] wrote:
>
>> Many mortgage companies/banks require auto-draft from checking accounts
>> for mortgage payments (mine does). I use auto-pay from my credit card
>> for many things, just too darn convenient to not do so. The only hassle
>> is to update all the auto-pay accounts every few years when the cc expires.
>
>Hmm, I've had many mortgages and never found one that had that
>requirement. ...

So it must not be real. How typical of you.

--
Best regards, SEE THE FAQ FOR CINGULAR WIRELESS AT
John Navas <http://en.wikibooks.org/wiki/Cingular_Wireless_FAQ>

---

› See More: American Consumer Satisfaction Index (ACSI) Report Evaluates ConsumerSatisfaction with Wireless Carriers. T-Mobile and Verizon top ratings, Cingularand Sprint tied at the bottom.

[POSTED TO alt.cellular.cingular - REPLY ON USENET PLEASE]

In <[email protected]> on Thu, 18 May 2006
14:44:14 -0500, "Thomas T. Veldhouse" <[email protected]> wrote:

>In alt.cellular.sprintpcs John Navas <[email protected]> wrote:
>>
>> Did I say that? No. But to be (even more) clear:
>>
>> * I don't put any stock in surveys and polls conducted by the entity
>> benefitting from the results.
>
>You are a naturally distrustful person apparently.

I think it's more a case of being realistic.

>It is entirely within the
>realm of reason that a company will conduct such "polling" or review so that
>they can maintain the quality of their product.

Bias is hard to avoid even when there isn't a vested interest in the outcome.

>It is a truly self-serving
>interest as to why they do it.

My own experience is that very few companies are that objective.

>In that case, the information is as accurate
>as they need it to be and would serve no purpose to have it skewed. I am not
>saying that is the case in general, but I do believe that is the reasoning
>behind Verizon's field trucks checking coverage and call quality.

Looks like a big budget ad campaign to me, which pretty much eviscerates the
case for objectivity.

>> * I don't put much stock in surveys and polls paid for by the entity
>> benefitting from the results.
>
>My comments above apply here as well.

Mine too.

>> * I take with a grain of salt surveys and polls with poor methodology; e.g.,
>> small sample size; self-selected (non-random) sample; non-uniform universe;
>> low response rate; no assessment of non-response bias.
>
>Your assumption is that a given test did not have such assessments, when in
>fact, all you really know is that you don't have that information.

That's the only safe assumption that can be made.

>> * I take note of the margin of error.
>
>As should everybody.

All too many don't.

>>>From a personal
>>>perspective, I know more people happy with Verizon than the rest.
>>
>> My own experience is that all major carriers are roughly comparable in terms
>> of customer satisfaction, that differences are relatively small.
>
>I would disagree there. ...

Fair enough.

>>>Sprint PCS
>>>has a few more holes in the system, probably to be expected on PCS (requires a
>>>denser tower distribution).
>>
>> In metro areas there is no significant technology difference (due to less than
>> maximum range spacing), and even in non-metro areas the difference, due to
>> shorter range from lower permitted maximum power for 1900 MHz, tends to be
>> relatively small:
>>
>> * Maximum power in the 800 band is 3 watts.
>> * Maximum power in the 1900 band is 2 watts.
>
>All cells that overlap cause increases in noise, and thus generally reduce the
>number of concurrent calls and total bandwidth available.

That only applies to CDMA -- TDMA (which includes GSM) uses dedicated time
slots.

>However CDMA has
>the ability to communicate with multiple towers simultaneously during a call,
>where GSM must hand-off between individual towers.

CDMA also requires hand-offs, just differently.

>CDMA is simply more
>efficient with bandwidth, and has greater error correction, both of which
>means more calls in similar RF conditions.

As I wrote, the difference (as measured in Erlangs) is relatively small. From
one of my past posts:

Unlike a TDMA/GSM network, a CDMA network is interference-limited
rather than bandwidth-limited, and the Erlang capacity is calculated
according to the probability of blocking by the network, i.e., the
probability that a new mobile is denied access to the network.

The problem is that CDMA Erlang numbers don't take into account the
quality/usability of the call. CDMA networks can be loaded to the
point where quality is truly horrible, whereas TDMA/GSM networks
guarantee bandwidth to every call. Real world RF interference
further muddies the water.

The result is that claimed Erlangs (call capacity metrics) for CDMA
networks tend to be unrealistic (some would say wildly unrealistic);
i.e., they are not directly comparable to the capacity of GSM
networks. When CDMA calls are limited to those with quality
comparable to GSM, then capacity is roughly comparable.

Contrary to the claims of CDMA boosters, there is no magic.(c)

Some time ago, Chris Pearson, Executive Vice President, 3G Americas, wrote:

Many CDMA operators are currently in the midst of deploying 1XRTT, AN
INTERIM STEP TOWARDS 3G that promises to use spectrum more
efficiently. Time will tell whether that is the truth but the fact is
that, based on best-case data from CDMA vendors, 1XRTT with EVRC
handles up to 156 Erlangs per sector. Bearing in mind that GSM with
AMR handles 142 Erlangs, it is a great stretch to argue that 1XRTT
has a major advantage over GSM. GSM operators also can deploy dynamic
frequency and channel allocation (DFCA), which assigns calls to
channels based on conditions such as signal and interference. With
AMR and DFCA, GSM can handle 170 Erlangs per sector - an improvement
on 1XRTT’s 156. [emphasis added]

In the near future, 1XRTT operators will probably be able to deploy a
technology called selective mode vocoder (SMV), which could provide
20% more capacity over EVRC. The catch is that SMV-like methods can
be applied to GSM to produce almost identical capacity gains. Thus,
while one technology may have slightly higher capacity gains at one
point in time, another technology is always preparing to leap-frog
over it.

>> Both GSM and CDMA have pros and cons; e.g.,
>>
>> * CDMA cell "breathing" can result in degraded quality and dropped calls.
>
>cell breathing is the result of decreased SNR. With GSM, you get whole
>channels.

Time slots.

>With CDMA it works more like IP packets.

CDMA is actually quite different from IP packets, which is why cell
"breathing" is an issue.

>Essentially, what this
>means is that with an decreased SNR amounts to a reduced radius of coverage
>with CDMA. THe SNR is caused by call volume.

Essentially correct.

>With GSM, you simply run out of
>channels and callers get "no service".

Again, time slots.

>The overall number of users possible
>on a CDMA system is larger than is possible on GSM due to the finer
>resolutions of data transfer capable with CDMA.

That's not the actual difference, but again, the difference in capacity is
relatively small, and irrelevant to subscribers.

>> * TDMA (GSM) can drop calls when moving to a cell running at full capacity.
>>
>> This of course varies from location to location, but carriers work hard to
>> minimize such problems, and both technologies work well in general.
>
>I indicate this above .. GSM and TDMA are simply out of channels. With CDMA,
>it is just a matter of noise (SNR). If there is a high enough SNR to send
>data at a fast enough rate to make a call, then you can connect.

Sure, but the CDMA connection can be horribly unusable, making the difference
moot. (I speak from considerable personal experience. Since network busy
isn't a significant problem, I'd personally rather have dedicated time slot,
but as always, YMMV.)

>>>Verizon offers CDMA, which makes better use of
>>>available bandwidth,
>>
>> The difference is small, and irrelevant from a subscriber standpoint.
>
>Care to back that up? ...

See above for the first part. The second part is patently clear.

>> As of a couple of years ago, Verizon actually had the least spectrum per
>> subscriber of the major carriers, and Cingular gained considerable spectrum
>> and favorable tower locations as a result of the merger with ATTWS, but much
>> has changed since then, and I know of no reliable public data on the current
>> situation.
>
>Indeed ... but Verizon has agreements with many others, thus increasing the
>effective capacity, which is all that matters to the user. Do you really care
>if you are on Verizon's or Alltel's network as long as all the features exist
>for you (that is a big if)?

Doesn't work that way -- roaming is only permitted in areas where there is no
native network coverage, not as a matter of capacity in native coverage areas.

>>>The big disclaimer I see with Verizon is cost ...
>>
>> Another is the lack of mobile device flexibility as compared to the open
>> SIM-based GSM standard, as well as more limited device configuration.
>
>I think that really has less to do with CDMA and more to do with hardware
>implementation choices.

[shrug]

>>>they are notoriously
>>>expensive, but now I see Cingular has headed in that direction as well.
>>
>> That's likely because Cingular has become more competitive with Verizon
>> (amount of corporate resources, economy of scale, number of subscribers,
>> spectrum, towers, etc.) as a result of the merger with ATTWS. (As the
>> smallest of the major carriers, T-Mobile is still buying market share.)
>
>Yes, they are the top dog by sheer numbers of customers. However, take a look
>at other things, like attrition of the market. Mergers have immediate effects
>on numbers, but equilibrium is where the numbers have true meaning.

Cingular is doing well by any reasonable measure.

>BTW ... you are clearly a Cingular supporter, absolutely no doubt judging by
>your comments here. I don't advise denying that in the future.

Nonsense. You only discredit your own argument by claiming that facts
contrary to your point of view much be biased. 'Those who have evidence will
present their evidence, whereas those who do not have evidence will attack the
man.' Not only that, but arguments over which is the "best" technology are
pretty pointless from a subscriber point of view. All that matters is what
works, and both technologies work quite well.

--
Best regards, SEE THE FAQ FOR CINGULAR WIRELESS AT
John Navas <http://en.wikibooks.org/wiki/Cingular_Wireless_FAQ>

[POSTED TO alt.cellular.cingular - REPLY ON USENET PLEASE]

In <[email protected]> on Thu, 18 May 2006
14:56:39 -0500, "Thomas T. Veldhouse" <[email protected]> wrote:

>In any event, as I
>have indicated over and over, you have not cited non-response bias, simply the
>potential for it to exist based on the response pattern.

That's sufficient. The burden of proof is on those conducting the survey.

--
Best regards, SEE THE FAQ FOR CINGULAR WIRELESS AT
John Navas <http://en.wikibooks.org/wiki/Cingular_Wireless_FAQ>

"John Navas" <[email protected]> wrote in message
news:[email protected]...
> [POSTED TO alt.cellular.cingular - REPLY ON USENET PLEASE]
>
> In <[email protected]> on Thu, 18 May 2006
> 14:56:39 -0500, "Thomas T. Veldhouse" <[email protected]> wrote:
>
>>In any event, as I
>>have indicated over and over, you have not cited non-response bias, simply
>>the
>>potential for it to exist based on the response pattern.
>
> That's sufficient. The burden of proof is on those conducting the survey.
>

And they don't claim any. So, according to your own logic there is none.
End of story.

"John Navas" <[email protected]> wrote in message
news:[email protected]...
> [POSTED TO alt.cellular.cingular - REPLY ON USENET PLEASE]
>
> In <[email protected]> on Wed, 17 May 2006 20:05:40
> -0700, SMS <[email protected]> wrote:
>
>>Scott wrote:
>>
>>> Rubbish- you simply can't stand that Cingular is being rated so poorly.
>>> And
>>> BTW- a 10% response rate is considered high.
>>
>>Actually it's considered extremely high, resulting in an extremely low
>>margin of error. ...
>
> You're either blinded by your bias or ignorant of statistics -- which is
> it?

Neither- you're the only one who can't see around here.

>
> NON-RESPONSE BIAS
> If you try to survey 100 people, and 40 of them don't respond, those 40
> could be different in some important way from the 60 who did respond.

Does that say "could be different"? So that infers that the 40 "could be
the same". That proves nothing except that you like to quote very poorly
written material.


> That's non-response bias - a problem often ignored in survey research.
> Non-response bias can be estimated by comparing data on the current
> sample
> with other data (e.g. from a Census) on the same population.
> <http://www.audiencedialogue.org/gloss-data.html>

Another Australian cite- how interesting that you can't find any reliable
sources.

>
> NONRESPONSE BIAS ANALYSES AT THE NATIONAL CENTER FOR EDUCATION
> STATISTICS
> In surveys with low response rates, non-response bias can be a major
> concern. While it is not always possible to measure the actual bias
> due to non-response, there are different approaches that help
> identify potential sources of non-response bias. In the National
> Center for Education Statistics (NCES), SURVEYS WITH A RESPONSE RATE
> LOWER THAN 70% MUST CONDUCT A NON-RESPONSE BIAS ANALYSIS. This paper
> discusses the different approaches to non-response bias analyses
> using examples from NCES. [emphasis added]
> <http://www.statcan.ca/bsolc/english/bsolc?catno=11-522-X20010016269>

Canadian this time- you're getting closer, but no more credible.

>
> The ACSI response rate, which results from the method used, is actually
> very
> low from a statistical standpoint (see above), which is why non-response
> bias
> is such a significant issue. To minimize non-response bias, responsible
> statisticians strive for response rates well in excess of 50%, but that
> takes
> follow-up with non-respondendents, which is time-consuming and expensive.
> See:

This one speaks of response rate, which you certainly did not plan on using.
I'm going to use this approach to survey Cingular- I'm going to ask thirty
users their opinion of the carrier. According to this, as long as more than
16 of them answer the question, I'll have a statistically valid sample and
it will represent the view of all Cingular subscribers.

Which is more accurate- an initial sample of 100,000 with a 10% take rate or
an initial sample of 10,000 with a 75% take rate (assuming both customer
bases are equal in size)?


>
> * Berg, N. Non-response bias, In Kempf-Leonard, K. (ed.), Encyclopedia
> of Social Measurement vol. 2, pp. 865-873, London, Academic Press.
>
> <http://www.utdallas.edu/~nberg/Berg_ARTICLES/BergNon-ResponseBiasMay2002.pdf>
>
> * Investigating non-response bias in a survey of disablement in the
> community: implications for survey methodology, A Tennant and EM
> Badley
> ARC Epidemiology Research Unit, University of Manchester, United
> Kingdom
> <http://jech.bmjjournals.com/cgi/content/abstract/45/3/247>
>
> * A confidence interval approach to investigating non-response bias and
> monitoring response to postal questionnaires, A Tennant and EM Badley
> ARC Epidemiology Research Unit, University of Manchester, United
> Kingdom
> <http://jech.bmjjournals.com/cgi/content/abstract/45/1/81>
>
> * National Health And Nutrition Examination Survey III,
> Accounting For Item Nonresponse Bias, Westat, Inc.
>
> <http://www.cdc.gov/nchs/data/nhanes/nhanes3/cdrom/NCHS/MANUALS/NR_BIAS.PDF>
>
> * Research Paper: Low Response Rates and Their Effects on Survey Results
>
> <http://www.sch.abs.gov.au/SCH%5CA1610103.NSF/0/3CE43BABF8BBF59DCA256B7C0001AEA4?OpenDocument=>
>
> * "A White Paper on Research Sampling" by Insight MAS
> <http://www.insightmas.com/documents/WhitePaper-Sampling.pdf>

A couple of these sites disclaim their ability to verify the accuracy of the
information posted. That certainly makes it believeable data.

Your isolation of this point in survey analysis must mean that you've
discovered the error of your other inaccurate opinions. And once again you
show that you are far from smart or all-knowing as you try to portray- all
you can do is Google other's work while having no practical experience of
your own. How pedestrian- this makes my 10 year old as smart as you, as he
can do the same thing.

"SMS" <[email protected]> wrote in message
news:[email protected]...
> Scott wrote:
>> "SMS" <[email protected]> wrote in message
>> news:[email protected]...
>>> Scott wrote:
>>>
>>>> Rubbish- you simply can't stand that Cingular is being rated so poorly.
>>>> And BTW- a 10% response rate is considered high.
>>> Actually it's considered extremely high, resulting in an extremely low
>>> margin of error.
>>
>> I'll have to do a little digging in the morning to see what the expected
>> take rate on our company survey is (done by an outside vendor)- I want to
>> say that 11 or 12% is a good day.
>
> Well I should clarify what I said. Surveying 10% of the total users is
> considered a very high rate, with a very low margin of error. But sending
> out surveys to 10% of the total users, and getting only a 10% response
> rate on those 10%, would be a low response rate. A reputable survey firm
> would not even report results in this case. I.e., in Consumer Report's
> wireless survey, they don't state the results if the number or responses
> is too low for an accurate result.

I checked- 60% sample of the target customer base (on average) with a 17%
take rate (on average). Sample is a gathering of account numbers not
contacted in the previous six months (which are totally random in regard to
geography, income, gender, account type) with nothing other than date of
interaction queried. Sample is compared to a seperate smaller sample (same
criteria as above) contacted by in-house employees (with exactly the same
survey) with an average monthly variance of .07% over the previous 24
months. Survey contacts are made daily (seven days a week) to eliminate the
possibility of missing short term issues that could affect customer
satisfaction. Tens of thousands of contacts a month on a monthly target base
of 1.5-2 million customers.

Now, Navas- I had to check on the sample size and take rate because I don't
have direct involvement in that. The rest of the information I didn't have
to check on, because the requirements and development of the survey
(including questions asked) and post-survey analysis of the data had to get
my sign-off before implementation and publication. Now- care to share your
experience?

"SMS" <[email protected]> wrote in message news:[email protected]...
>> My point though is that they can have a glitch, or a billing error and bill
>> you for FAR more than you owe. If you are on auto-pay, that money is taken
>> from you and you fight to get it back, usually with a significant wait that
>> amounts to a credit on your next bill.
>
> No, the credit card company credits you immediately while they are
> investigating. However if you're wrong, you pay interest on the charge
> when it's confirmed.

You're right about the immediate credit. As far as finance charges, I think
it varies by company. I recently disputed two charges on a Citibank
credit card. After two months, one charge was decided in my favor, the
second was upheld as valid, but they did not charge a finance charge on it.
OTOH, I always pay my credit card bills in full each month (minus any disputed
amounts), so perhaps that made a difference.

--
John Richards

"John Navas" <[email protected]> wrote in message news:[email protected]...
> [POSTED TO alt.cellular.cingular - REPLY ON USENET PLEASE]
>
> In <[email protected]> on Thu, 18 May 2006 15:33:24
> -0700, SMS <[email protected]> wrote:
>
>>[email protected] wrote:
>>
>>> Many mortgage companies/banks require auto-draft from checking accounts
>>> for mortgage payments (mine does). I use auto-pay from my credit card
>>> for many things, just too darn convenient to not do so. The only hassle
>>> is to update all the auto-pay accounts every few years when the cc expires.
>>
>>Hmm, I've had many mortgages and never found one that had that
>>requirement. ...
>
> So it must not be real. How typical of you.

It must not be very common. Over the past 40 years I've had
mortgage loans from at least 10 different companies, and was never
required to have auto-draft from checking. Perhaps if one's
credit history was bad enough, it might be a requirement.

--
John Richards

Scott wrote:

>> NONRESPONSE BIAS ANALYSES AT THE NATIONAL CENTER FOR EDUCATION
>> STATISTICS
>> In surveys with low response rates, non-response bias can be a major
>> concern. While it is not always possible to measure the actual bias
>> due to non-response, there are different approaches that help
>> identify potential sources of non-response bias. In the National
>> Center for Education Statistics (NCES), SURVEYS WITH A RESPONSE RATE
>> LOWER THAN 70% MUST CONDUCT A NON-RESPONSE BIAS ANALYSIS. This paper
>> discusses the different approaches to non-response bias analyses
>> using examples from NCES. [emphasis added]
>> <http://www.statcan.ca/bsolc/english/bsolc?catno=11-522-X20010016269>
>
> Canadian this time- you're getting closer, but no more credible.

Look at the first line, "In surveys with low response rates." That's the
key point that Navas is ignoring. Once you have a high response rate,
non-response bias isn't an issue, unless you're doing some sort of
survey where there is a reason that one group would be much less likely
to respond. In terms of wireless, the surveys by Consumer Reports, JD
Power, and others, aren't asking people "which carrier do you think is
the best?" they are surveying individuals regarding their experience
with their own carrier. Thus, it's not a valid complaint to claim that
maybe the reason that Cingular always does so poorly is because only
people that are unhappy with their wireless service choose to respond
because as long as the response rate for each carrier, in each region,
is sufficient, the results are valid. Occasionally CR will leave out
ratings for a carrier in a specific region if they did not receive
enough responses. No one takes Navas's criticism of the CR surveys
seriously, it's simply sour grapes.

John Richards wrote:

> It must not be very common. Over the past 40 years I've had
> mortgage loans from at least 10 different companies, and was never
> required to have auto-draft from checking. Perhaps if one's
> credit history was bad enough, it might be a requirement.

That's probably what they do, if your credit score is below a certain
level, or maybe if the LTV is below a certain ratio, then they require
auto-pay. I know that some banks will jack up the base rate, then offer
a discount for auto-pay, but the places that already have the lowest
rates don't do this.

If you look at the wholesale rate sheets from the wholesale lenders,
there is no discount for auto-pay, it's strictly a way for banks to try
to get new accounts. On the wholesale rate sheet there are several
discounts that a bank or broker can get (and choose to pass on to you):

1. Loan to value ratio above a certain percentage
2. High credit score (usually above 740)
3. Impound-escrow for taxes and/or property insurance

I always use the FAXBACK service of the wholesale lenders to check on
what discounts are currently being offered, as if you know about them,
you're in a better bargaining position.

Thomas T. Veldhouse wrote:
> In alt.cellular.sprintpcs John Navas <[email protected]> wrote:
>> Did I say that? No. But to be (even more) clear:
>>
>> * I don't put any stock in surveys and polls conducted by the entity
>> benefitting from the results.
>>
>
> You are a naturally distrustful person apparently. It is entirely within the
> realm of reason that a company will conduct such "polling" or review so that
> they can maintain the quality of their product. It is a truly self-serving
> interest as to why they do it. In that case, the information is as accurate
> as they need it to be and would serve no purpose to have it skewed. I am not
> saying that is the case in general, but I do believe that is the reasoning
> behind Verizon's field trucks checking coverage and call quality.

This is all true, but the reason that they make a big deal about it is
because the results turned out favorably for them. If they had done
poorly in coverage, then they would obviously not used the results as
the basis for an advertising campaign. Clearly the other carriers do
coverage checking too, but they don't want to use the results of their
tests in an ad campaign, for obvious reasons. Verizon was emboldened to
use their test results by the corroboration from independent sources,
since they now can point to the other surveys as proof that their tests
were fair.

What Cingular is doing with Telephia may trick some naive buyers, but it
also has the opposite effect on buyers that actually dissect ads. It's
always interesting when a company tries to focus on one unprovable and
irrational premise as a basis for promoting their product, and to look
at the contortions they go through, and the weasel words they use.
Sprint's "The Largest All Digital Network" campaign was beautiful, as
the effectively eliminated Verizon and Cingular from the competition.
Cingular's attempt to equate the number of dropped calls with the best
network is so transparently false, that you have to wonder at the
desperation that made them say this. Yet many people don't realize that
with Cingular, the fact that you can't even begin a call in a lot of
areas that have have Verizon or Sprint coverage, means that you'll have
fewer calls to drop in the first place. Anyone that reads the inpdendent
surveys realizes this, but the informed consumer is not Cingular's
target market.

In alt.cellular.sprintpcs John Richards <[email protected]> wrote:
>
> It must not be very common. Over the past 40 years I've had
> mortgage loans from at least 10 different companies, and was never
> required to have auto-draft from checking. Perhaps if one's
> credit history was bad enough, it might be a requirement.
>

I am on company five and I have never seen a mortgage company require autopay
either. A few offer it. I prefer to manually setup payment each month ... no
surprises that way.

--
Thomas T. Veldhouse
Key Fingerprint: 2DB9 813F F510 82C2 E1AE 34D0 D69D 1EDC D5EC AED1

In alt.cellular.sprintpcs John Navas <[email protected]> wrote:
>>
>>You are a naturally distrustful person apparently.
>
> I think it's more a case of being realistic.
>

Reading your comments below ... you are clearly the real Thomas and I am an
imposter ... I do not hold the doubt and cynacism so close to my heart.

>>It is entirely within the
>>realm of reason that a company will conduct such "polling" or review so that
>>they can maintain the quality of their product.
>
> Bias is hard to avoid even when there isn't a vested interest in the outcome.
>

You have aggressively post editted my quote and removed all references to the
fact that I am talking about Verizon here. Verizon is doing QC work with
those mobile vans. The data has its marketting uses obviously, and they are
using it. They do not hide the fact that they are the company gathering the
information. As an informed consumer, I am surprised you take such a negative
opinion of it.

>>It is a truly self-serving
>>interest as to why they do it.
>
> My own experience is that very few companies are that objective.
>

It is their own QC data ... there is no point in doing QC work if you don't do
it objectively.

>>In that case, the information is as accurate
>>as they need it to be and would serve no purpose to have it skewed. I am not
>>saying that is the case in general, but I do believe that is the reasoning
>>behind Verizon's field trucks checking coverage and call quality.
>
> Looks like a big budget ad campaign to me, which pretty much eviscerates the
> case for objectivity.

Cynic. Cynic ... cynic!

>
>>> * I take with a grain of salt surveys and polls with poor methodology; e.g.,
>>> small sample size; self-selected (non-random) sample; non-uniform universe;
>>> low response rate; no assessment of non-response bias.
>>
>>Your assumption is that a given test did not have such assessments, when in
>>fact, all you really know is that you don't have that information.
>
> That's the only safe assumption that can be made.
>

Cynic! Why don't you try and find out if you want to know?

>>> * I take note of the margin of error.
>>
>>As should everybody.
>
> All too many don't.

Again, that is not a problem with the statistics, that is a problem with the
people reading them. Period. It is NOT a problem with the company quoting
them.

>>All cells that overlap cause increases in noise, and thus generally reduce the
>>number of concurrent calls and total bandwidth available.
>
> That only applies to CDMA -- TDMA (which includes GSM) uses dedicated time
> slots.
>

Of course they have dedicated time slots (that is what T stands for in TDMA
.... GSM just has more of them and they number of slots varies during a call).

The only way to overlap TDMA/GSM zones are to put them on different bands.
CDMA does not have to worry about it. The interference caused by overlapping
GSM/TDMA is absolute, so deployment is much more problematic.

>>However CDMA has
>>the ability to communicate with multiple towers simultaneously during a call,
>>where GSM must hand-off between individual towers.
>
> CDMA also requires hand-offs, just differently.

A phone can talk with two or three towers at the same time ... where one tower
is dominant. Then, as the phone moves, tower dominance moves and eventually
one tower ceases to communicate with the phone, at some point another tower is
reached ... and the phone begins talking with it as well ... until the whole
cycle starts again .. there is no *direct* handoff per se ... unless one were
to cross a carrier boundry (not sure if that is even supported these days).

>
>>CDMA is simply more
>>efficient with bandwidth, and has greater error correction, both of which
>>means more calls in similar RF conditions.
>
> As I wrote, the difference (as measured in Erlangs) is relatively small. From
> one of my past posts:
>
<snip<snip<snip<snip<snip<snip<snip<snip<snip>

I challenge you to get real statistics on the call density of urban towers
using CDMA. I have seen it widely reported that CDMA towers are getting
higher call density. Verizon is known to limit the number of calls to a cap
to avoid the severe call degradation that the snipped text mentions. At times
though, it is better to make the call with a little lower voice quality than
it is to not be able to make the call at all. Think about the performance of
phones in Manhattan on 9/11.

> Some time ago, Chris Pearson, Executive Vice President, 3G Americas, wrote:
>
> Many CDMA operators are currently in the midst of deploying 1XRTT, AN
> INTERIM STEP TOWARDS 3G that promises to use spectrum more
> efficiently. Time will tell whether that is the truth but the fact is
> that, based on best-case data from CDMA vendors, 1XRTT with EVRC
> handles up to 156 Erlangs per sector. Bearing in mind that GSM with
> AMR handles 142 Erlangs, it is a great stretch to argue that 1XRTT
> has a major advantage over GSM. GSM operators also can deploy dynamic
> frequency and channel allocation (DFCA), which assigns calls to
> channels based on conditions such as signal and interference. With
> AMR and DFCA, GSM can handle 170 Erlangs per sector - an improvement
> on 1XRTT?s 156. [emphasis added]
>

1xRtt has been out for many years now. Now, what are the field results ...
today, not at the time of that old text.

> In the near future, 1XRTT operators will probably be able to deploy a
> technology called selective mode vocoder (SMV), which could provide
> 20% more capacity over EVRC. The catch is that SMV-like methods can
> be applied to GSM to produce almost identical capacity gains. Thus,
> while one technology may have slightly higher capacity gains at one
> point in time, another technology is always preparing to leap-frog
> over it.
>

Indeed ... so what are the real world results again?

>>> Both GSM and CDMA have pros and cons; e.g.,
>>>
>>> * CDMA cell "breathing" can result in degraded quality and dropped calls.
>>
>>cell breathing is the result of decreased SNR. With GSM, you get whole
>>channels.
>
> Time slots.

Pardon me ... that is what I am referring too. Actually, CDMA allocates
channels, but it can be several thousand in a minute where CDMA spreads its
packets accross the entire available spectrum allocated at a given location.

>
>>With CDMA it works more like IP packets.
>
> CDMA is actually quite different from IP packets, which is why cell
> "breathing" is an issue.

I understand perfectly why breathing occurs. The analogy is best described as
noise ... as SNR drops the cell shrinks. The only way to make a call in this
case is to increase signal (power) or to get closer (effectively increasing
power). The IP packets I refer to have nothing to do with breathing. Voice
is broken up into small "packets" and then the phone looks for an available
channel and sends the "packet". The "breathing" you refer to is a result of
spread spectrum and not the protocol directly. The protocol is only relavent
in the fact that is is capable of working in a spread spectrum scenario ... it
should be possible for the protocol to work fine over a different air
interface, such as the timeslots used with GSM and TDMA.

>
>>Essentially, what this
>>means is that with an decreased SNR amounts to a reduced radius of coverage
>>with CDMA. THe SNR is caused by call volume.
>
> Essentially correct.
>
>>With GSM, you simply run out of
>>channels and callers get "no service".
>
> Again, time slots.

Agreed.

>
>>The overall number of users possible
>>on a CDMA system is larger than is possible on GSM due to the finer
>>resolutions of data transfer capable with CDMA.
>
> That's not the actual difference, but again, the difference in capacity is
> relatively small, and irrelevant to subscribers.

Actually, it is the difference. The "packet" of data sent in CDMA is smaller
than that in GSM, thus, one can approach the theoretical limit more closely.

>
> Sure, but the CDMA connection can be horribly unusable, making the difference
> moot. (I speak from considerable personal experience. Since network busy
> isn't a significant problem, I'd personally rather have dedicated time slot,
> but as always, YMMV.)
>

Yes, but the horribly unusable really occurs after the same situation applied
to GSM would result in no call at all. My point is that CDMA can get higher
capacity.

>>>>Verizon offers CDMA, which makes better use of
>>>>available bandwidth,
>>>
>>> The difference is small, and irrelevant from a subscriber standpoint.
>>
>>Care to back that up? ...
>
> See above for the first part. The second part is patently clear.

Again, what you posted is several years old. Still, having said that I would
like to see real world statistics as to real capacities at real sites. I have
read over and over that CDMA has been offering better capacity, which is major
(but not only) reason why Vodofone(?) has not forced Verizon to convert to GSM
like the rest of its companies use.

>
>>> As of a couple of years ago, Verizon actually had the least spectrum per
>>> subscriber of the major carriers, and Cingular gained considerable spectrum
>>> and favorable tower locations as a result of the merger with ATTWS, but much
>>> has changed since then, and I know of no reliable public data on the current
>>> situation.
>>
>>Indeed ... but Verizon has agreements with many others, thus increasing the
>>effective capacity, which is all that matters to the user. Do you really care
>>if you are on Verizon's or Alltel's network as long as all the features exist
>>for you (that is a big if)?
>
> Doesn't work that way -- roaming is only permitted in areas where there is no
> native network coverage, not as a matter of capacity in native coverage areas.
>

Sure it does. If you ask a user, are you happy and that user answers yes,
then they are doing something correctly. If they are doing it by making an
agreement with local provider XYZ to offer seamless coverage ... then it is
effective.

>>>>The big disclaimer I see with Verizon is cost ...
>>>
>>> Another is the lack of mobile device flexibility as compared to the open
>>> SIM-based GSM standard, as well as more limited device configuration.
>>
>>I think that really has less to do with CDMA and more to do with hardware
>>implementation choices.
>
> [shrug]
>

Shrug if you like. GSM doesn't mean SIM. GSM CAN be implemented without SIMs
and CDMA CAN be implemented with SIMs. They are independent technologies that
are traditionally used together.

>>Yes, they are the top dog by sheer numbers of customers. However, take a look
>>at other things, like attrition of the market. Mergers have immediate effects
>>on numbers, but equilibrium is where the numbers have true meaning.
>
> Cingular is doing well by any reasonable measure.
>
>>BTW ... you are clearly a Cingular supporter, absolutely no doubt judging by
>>your comments here. I don't advise denying that in the future.
>
> Nonsense. You only discredit your own argument by claiming that facts
> contrary to your point of view much be biased. 'Those who have evidence will
> present their evidence, whereas those who do not have evidence will attack the
> man.' Not only that, but arguments over which is the "best" technology are
> pretty pointless from a subscriber point of view. All that matters is what
> works, and both technologies work quite well.
>

Hehe ... I will leave that to the audience .. but I have watched you respond
here over the last couple of days, to me and to others, and it is nearly
always in support of Cingular in an argument against them ... except perhaps
for their claim of the lowest number of dropped calls.

--
Thomas T. Veldhouse
Key Fingerprint: 2DB9 813F F510 82C2 E1AE 34D0 D69D 1EDC D5EC AED1

In alt.cellular.sprintpcs John Navas <[email protected]> wrote:
> [POSTED TO alt.cellular.cingular - REPLY ON USENET PLEASE]
>
> In <[email protected]> on Thu, 18 May 2006
> 14:56:39 -0500, "Thomas T. Veldhouse" <[email protected]> wrote:
>
>>In any event, as I
>>have indicated over and over, you have not cited non-response bias, simply the
>>potential for it to exist based on the response pattern.
>
> That's sufficient. The burden of proof is on those conducting the survey.
>

Actually, it is not. It is YOU with the doubt. Why don't YOU request the
data, if they refuse to supply it, then yes, you might surmise they don't want
to tell you because the bias is high, but right now, you have inferred data
that you don't have and can not reasonably say is likely the case.

--
Thomas T. Veldhouse
Key Fingerprint: 2DB9 813F F510 82C2 E1AE 34D0 D69D 1EDC D5EC AED1

[POSTED TO alt.cellular.cingular - REPLY ON USENET PLEASE]

In <[email protected]> on Fri, 19 May 2006 08:05:43 -0500,
"Thomas T. Veldhouse" <[email protected]> wrote:

>In alt.cellular.sprintpcs John Navas <[email protected]> wrote:
>>
>> In <[email protected]> on Thu, 18 May 2006
>> 14:56:39 -0500, "Thomas T. Veldhouse" <[email protected]> wrote:
>>
>>>In any event, as I
>>>have indicated over and over, you have not cited non-response bias, simply the
>>>potential for it to exist based on the response pattern.
>>
>> That's sufficient. The burden of proof is on those conducting the survey.
>
>Actually, it is not. ...

We'll just have to agree to disagree. Have a nice day.

--
Best regards,
John Navas <http://navasgrp.home.att.net/>

"Usenet is like a herd of performing elephants with diarrhea - massive,
difficult to redirect, awe inspiring, entertaining, and a source of mind
boggling amounts of excrement when you least expect it." --Gene Spafford