Is the Tucson all that much better than the Santa Fe ?
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
Christopher Wong wrote:
> You are comparing apples and oranges. The JD Power IQS is NOT a reliability
> survey. It measures quality issues over a 90 day period
Right, and by the same logic, Consumer Reports is not really a
reliability survey either -- it's a "historical reliability survey."
Here's the problem: CR sends out surveys in the spring and reports the
results in November. This means the CR data you read today reflects
what owners thought in 2005. Any major changes in quality occurring in
2006 is missed by current CR data and won't be published until this
November.
For car companies whose quality and reliability patterns are relatively
static, CR ratings are fine. But Hyundai is improving quality by leaps
and bounds. CR's methodology is too slow to give an accurate picture of
what's going on ^today.^
With JD Power's 90-day survey you at least get a timely snapshot of
recent owner experiences. If they have to take their cars back to the
dealer frequently due to quality issues, it's going to show up in
Power's survey.
> The reason for the reliability disparity is quite simple. While Hyundai has
> made remarkable progress in reliability, they may not be as consistent
> across all models or when first starting out. The Tucson is still very new,
For Hyundai, IMHO, I think you have it backwards. They are making
strides in quality very quickly, and the trend seems to be the newer
the design the better the build. This does not surprise me, the Koreans
are world leaders in computerized design, manufacturing and automation.
This may be why the newer models are doing so well in JD Power (fresher
data) and not Consumer Reports (last years data). In fact, CR wrote
that the predicted reliability of the 2006 Tucson will be "worse than
average." From what we know today, that prediction has proven to be
dead wrong.
> You are comparing apples and oranges. The JD Power IQS is NOT a reliability
> survey. It measures quality issues over a 90 day period
Right, and by the same logic, Consumer Reports is not really a
reliability survey either -- it's a "historical reliability survey."
Here's the problem: CR sends out surveys in the spring and reports the
results in November. This means the CR data you read today reflects
what owners thought in 2005. Any major changes in quality occurring in
2006 is missed by current CR data and won't be published until this
November.
For car companies whose quality and reliability patterns are relatively
static, CR ratings are fine. But Hyundai is improving quality by leaps
and bounds. CR's methodology is too slow to give an accurate picture of
what's going on ^today.^
With JD Power's 90-day survey you at least get a timely snapshot of
recent owner experiences. If they have to take their cars back to the
dealer frequently due to quality issues, it's going to show up in
Power's survey.
> The reason for the reliability disparity is quite simple. While Hyundai has
> made remarkable progress in reliability, they may not be as consistent
> across all models or when first starting out. The Tucson is still very new,
For Hyundai, IMHO, I think you have it backwards. They are making
strides in quality very quickly, and the trend seems to be the newer
the design the better the build. This does not surprise me, the Koreans
are world leaders in computerized design, manufacturing and automation.
This may be why the newer models are doing so well in JD Power (fresher
data) and not Consumer Reports (last years data). In fact, CR wrote
that the predicted reliability of the 2006 Tucson will be "worse than
average." From what we know today, that prediction has proven to be
dead wrong.
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
> Christopher Wong wrote:
>> The reason for the reliability disparity is quite simple. While Hyundai has
>> made remarkable progress in reliability, they may not be as consistent
>> across all models or when first starting out. The Tucson is still very new,
>
> For Hyundai, IMHO, I think you have it backwards. They are making
> strides in quality very quickly, and the trend seems to be the newer
> the design the better the build. This does not surprise me, the Koreans
> are world leaders in computerized design, manufacturing and automation.
> This may be why the newer models are doing so well in JD Power (fresher
> data) and not Consumer Reports (last years data). In fact, CR wrote
> that the predicted reliability of the 2006 Tucson will be "worse than
> average." From what we know today, that prediction has proven to be
> dead wrong.
I say you are way too optimistic -- you are predicting the future -- based
on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
can't extrapolate reliability from this study. Nor should it suggest that
the current Tucson should improve significantly: why do you assume it
improved its IQS score? It has always done well. According to Edmunds, the
2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
on the IQS. So naively going by the unchanged IQS score, we can assume that
the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
category, tying with the Kia Sportage and Saturn Vue.
While I am optimistic about Hyundai's quality in general, it sure looks like
the Tucson is a dud. This is the first Hyundai in years that scored this
horribly. The current Sonata debuted on the charts with an above-average
rating. The Santa Fe was always respectable across all years. The old Accent
and Elantra never scored this bad out the door. My fear is that there is
something badly wrong with the Tucson. You seem to be expecting miracles to
happen with the Tucson, but Hyundai simply does not have that kind of track
record. The Sonata is reliable now, but when it was introduced in 2000 it
spent the first 2 model years with below-average reliability. There is
absolutely no proof that Hyundai turned around the Tucson in just one year.
Chris
> Christopher Wong wrote:
>> The reason for the reliability disparity is quite simple. While Hyundai has
>> made remarkable progress in reliability, they may not be as consistent
>> across all models or when first starting out. The Tucson is still very new,
>
> For Hyundai, IMHO, I think you have it backwards. They are making
> strides in quality very quickly, and the trend seems to be the newer
> the design the better the build. This does not surprise me, the Koreans
> are world leaders in computerized design, manufacturing and automation.
> This may be why the newer models are doing so well in JD Power (fresher
> data) and not Consumer Reports (last years data). In fact, CR wrote
> that the predicted reliability of the 2006 Tucson will be "worse than
> average." From what we know today, that prediction has proven to be
> dead wrong.
I say you are way too optimistic -- you are predicting the future -- based
on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
can't extrapolate reliability from this study. Nor should it suggest that
the current Tucson should improve significantly: why do you assume it
improved its IQS score? It has always done well. According to Edmunds, the
2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
on the IQS. So naively going by the unchanged IQS score, we can assume that
the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
category, tying with the Kia Sportage and Saturn Vue.
While I am optimistic about Hyundai's quality in general, it sure looks like
the Tucson is a dud. This is the first Hyundai in years that scored this
horribly. The current Sonata debuted on the charts with an above-average
rating. The Santa Fe was always respectable across all years. The old Accent
and Elantra never scored this bad out the door. My fear is that there is
something badly wrong with the Tucson. You seem to be expecting miracles to
happen with the Tucson, but Hyundai simply does not have that kind of track
record. The Sonata is reliable now, but when it was introduced in 2000 it
spent the first 2 model years with below-average reliability. There is
absolutely no proof that Hyundai turned around the Tucson in just one year.
Chris
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
> Christopher Wong wrote:
>> The reason for the reliability disparity is quite simple. While Hyundai has
>> made remarkable progress in reliability, they may not be as consistent
>> across all models or when first starting out. The Tucson is still very new,
>
> For Hyundai, IMHO, I think you have it backwards. They are making
> strides in quality very quickly, and the trend seems to be the newer
> the design the better the build. This does not surprise me, the Koreans
> are world leaders in computerized design, manufacturing and automation.
> This may be why the newer models are doing so well in JD Power (fresher
> data) and not Consumer Reports (last years data). In fact, CR wrote
> that the predicted reliability of the 2006 Tucson will be "worse than
> average." From what we know today, that prediction has proven to be
> dead wrong.
I say you are way too optimistic -- you are predicting the future -- based
on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
can't extrapolate reliability from this study. Nor should it suggest that
the current Tucson should improve significantly: why do you assume it
improved its IQS score? It has always done well. According to Edmunds, the
2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
on the IQS. So naively going by the unchanged IQS score, we can assume that
the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
category, tying with the Kia Sportage and Saturn Vue.
While I am optimistic about Hyundai's quality in general, it sure looks like
the Tucson is a dud. This is the first Hyundai in years that scored this
horribly. The current Sonata debuted on the charts with an above-average
rating. The Santa Fe was always respectable across all years. The old Accent
and Elantra never scored this bad out the door. My fear is that there is
something badly wrong with the Tucson. You seem to be expecting miracles to
happen with the Tucson, but Hyundai simply does not have that kind of track
record. The Sonata is reliable now, but when it was introduced in 2000 it
spent the first 2 model years with below-average reliability. There is
absolutely no proof that Hyundai turned around the Tucson in just one year.
Chris
> Christopher Wong wrote:
>> The reason for the reliability disparity is quite simple. While Hyundai has
>> made remarkable progress in reliability, they may not be as consistent
>> across all models or when first starting out. The Tucson is still very new,
>
> For Hyundai, IMHO, I think you have it backwards. They are making
> strides in quality very quickly, and the trend seems to be the newer
> the design the better the build. This does not surprise me, the Koreans
> are world leaders in computerized design, manufacturing and automation.
> This may be why the newer models are doing so well in JD Power (fresher
> data) and not Consumer Reports (last years data). In fact, CR wrote
> that the predicted reliability of the 2006 Tucson will be "worse than
> average." From what we know today, that prediction has proven to be
> dead wrong.
I say you are way too optimistic -- you are predicting the future -- based
on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
can't extrapolate reliability from this study. Nor should it suggest that
the current Tucson should improve significantly: why do you assume it
improved its IQS score? It has always done well. According to Edmunds, the
2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
on the IQS. So naively going by the unchanged IQS score, we can assume that
the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
category, tying with the Kia Sportage and Saturn Vue.
While I am optimistic about Hyundai's quality in general, it sure looks like
the Tucson is a dud. This is the first Hyundai in years that scored this
horribly. The current Sonata debuted on the charts with an above-average
rating. The Santa Fe was always respectable across all years. The old Accent
and Elantra never scored this bad out the door. My fear is that there is
something badly wrong with the Tucson. You seem to be expecting miracles to
happen with the Tucson, but Hyundai simply does not have that kind of track
record. The Sonata is reliable now, but when it was introduced in 2000 it
spent the first 2 model years with below-average reliability. There is
absolutely no proof that Hyundai turned around the Tucson in just one year.
Chris
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
> Christopher Wong wrote:
>> The reason for the reliability disparity is quite simple. While Hyundai has
>> made remarkable progress in reliability, they may not be as consistent
>> across all models or when first starting out. The Tucson is still very new,
>
> For Hyundai, IMHO, I think you have it backwards. They are making
> strides in quality very quickly, and the trend seems to be the newer
> the design the better the build. This does not surprise me, the Koreans
> are world leaders in computerized design, manufacturing and automation.
> This may be why the newer models are doing so well in JD Power (fresher
> data) and not Consumer Reports (last years data). In fact, CR wrote
> that the predicted reliability of the 2006 Tucson will be "worse than
> average." From what we know today, that prediction has proven to be
> dead wrong.
I say you are way too optimistic -- you are predicting the future -- based
on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
can't extrapolate reliability from this study. Nor should it suggest that
the current Tucson should improve significantly: why do you assume it
improved its IQS score? It has always done well. According to Edmunds, the
2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
on the IQS. So naively going by the unchanged IQS score, we can assume that
the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
category, tying with the Kia Sportage and Saturn Vue.
While I am optimistic about Hyundai's quality in general, it sure looks like
the Tucson is a dud. This is the first Hyundai in years that scored this
horribly. The current Sonata debuted on the charts with an above-average
rating. The Santa Fe was always respectable across all years. The old Accent
and Elantra never scored this bad out the door. My fear is that there is
something badly wrong with the Tucson. You seem to be expecting miracles to
happen with the Tucson, but Hyundai simply does not have that kind of track
record. The Sonata is reliable now, but when it was introduced in 2000 it
spent the first 2 model years with below-average reliability. There is
absolutely no proof that Hyundai turned around the Tucson in just one year.
Chris
> Christopher Wong wrote:
>> The reason for the reliability disparity is quite simple. While Hyundai has
>> made remarkable progress in reliability, they may not be as consistent
>> across all models or when first starting out. The Tucson is still very new,
>
> For Hyundai, IMHO, I think you have it backwards. They are making
> strides in quality very quickly, and the trend seems to be the newer
> the design the better the build. This does not surprise me, the Koreans
> are world leaders in computerized design, manufacturing and automation.
> This may be why the newer models are doing so well in JD Power (fresher
> data) and not Consumer Reports (last years data). In fact, CR wrote
> that the predicted reliability of the 2006 Tucson will be "worse than
> average." From what we know today, that prediction has proven to be
> dead wrong.
I say you are way too optimistic -- you are predicting the future -- based
on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
can't extrapolate reliability from this study. Nor should it suggest that
the current Tucson should improve significantly: why do you assume it
improved its IQS score? It has always done well. According to Edmunds, the
2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
on the IQS. So naively going by the unchanged IQS score, we can assume that
the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
category, tying with the Kia Sportage and Saturn Vue.
While I am optimistic about Hyundai's quality in general, it sure looks like
the Tucson is a dud. This is the first Hyundai in years that scored this
horribly. The current Sonata debuted on the charts with an above-average
rating. The Santa Fe was always respectable across all years. The old Accent
and Elantra never scored this bad out the door. My fear is that there is
something badly wrong with the Tucson. You seem to be expecting miracles to
happen with the Tucson, but Hyundai simply does not have that kind of track
record. The Sonata is reliable now, but when it was introduced in 2000 it
spent the first 2 model years with below-average reliability. There is
absolutely no proof that Hyundai turned around the Tucson in just one year.
Chris
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
Christopher Wong wrote:
> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>> Christopher Wong wrote:
>>> The reason for the reliability disparity is quite simple. While Hyundai has
>>> made remarkable progress in reliability, they may not be as consistent
>>> across all models or when first starting out. The Tucson is still very new,
>> For Hyundai, IMHO, I think you have it backwards. They are making
>> strides in quality very quickly, and the trend seems to be the newer
>> the design the better the build. This does not surprise me, the Koreans
>> are world leaders in computerized design, manufacturing and automation.
>> This may be why the newer models are doing so well in JD Power (fresher
>> data) and not Consumer Reports (last years data). In fact, CR wrote
>> that the predicted reliability of the 2006 Tucson will be "worse than
>> average." From what we know today, that prediction has proven to be
>> dead wrong.
>
> I say you are way too optimistic -- you are predicting the future -- based
> on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
> can't extrapolate reliability from this study. Nor should it suggest that
> the current Tucson should improve significantly: why do you assume it
> improved its IQS score? It has always done well. According to Edmunds, the
> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
> on the IQS. So naively going by the unchanged IQS score, we can assume that
> the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
> the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
> category, tying with the Kia Sportage and Saturn Vue.
Consumer Reports, the people that put the Toyota Corolla and the Geo
Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
the early 90s.
> While I am optimistic about Hyundai's quality in general, it sure looks like
> the Tucson is a dud. This is the first Hyundai in years that scored this
> horribly. The current Sonata debuted on the charts with an above-average
> rating. The Santa Fe was always respectable across all years. The old Accent
> and Elantra never scored this bad out the door. My fear is that there is
> something badly wrong with the Tucson. You seem to be expecting miracles to
> happen with the Tucson, but Hyundai simply does not have that kind of track
> record. The Sonata is reliable now, but when it was introduced in 2000 it
> spent the first 2 model years with below-average reliability. There is
> absolutely no proof that Hyundai turned around the Tucson in just one year.
The Sonata was introduced long before 2000... I want to say 1989 or
1990. The Sonata V6 was introduced somewhere in the mid 90s.
Where exactly does CR get its statistics from? "Thin air" seems a
possible answer.
JS
> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>> Christopher Wong wrote:
>>> The reason for the reliability disparity is quite simple. While Hyundai has
>>> made remarkable progress in reliability, they may not be as consistent
>>> across all models or when first starting out. The Tucson is still very new,
>> For Hyundai, IMHO, I think you have it backwards. They are making
>> strides in quality very quickly, and the trend seems to be the newer
>> the design the better the build. This does not surprise me, the Koreans
>> are world leaders in computerized design, manufacturing and automation.
>> This may be why the newer models are doing so well in JD Power (fresher
>> data) and not Consumer Reports (last years data). In fact, CR wrote
>> that the predicted reliability of the 2006 Tucson will be "worse than
>> average." From what we know today, that prediction has proven to be
>> dead wrong.
>
> I say you are way too optimistic -- you are predicting the future -- based
> on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
> can't extrapolate reliability from this study. Nor should it suggest that
> the current Tucson should improve significantly: why do you assume it
> improved its IQS score? It has always done well. According to Edmunds, the
> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
> on the IQS. So naively going by the unchanged IQS score, we can assume that
> the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
> the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
> category, tying with the Kia Sportage and Saturn Vue.
Consumer Reports, the people that put the Toyota Corolla and the Geo
Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
the early 90s.
> While I am optimistic about Hyundai's quality in general, it sure looks like
> the Tucson is a dud. This is the first Hyundai in years that scored this
> horribly. The current Sonata debuted on the charts with an above-average
> rating. The Santa Fe was always respectable across all years. The old Accent
> and Elantra never scored this bad out the door. My fear is that there is
> something badly wrong with the Tucson. You seem to be expecting miracles to
> happen with the Tucson, but Hyundai simply does not have that kind of track
> record. The Sonata is reliable now, but when it was introduced in 2000 it
> spent the first 2 model years with below-average reliability. There is
> absolutely no proof that Hyundai turned around the Tucson in just one year.
The Sonata was introduced long before 2000... I want to say 1989 or
1990. The Sonata V6 was introduced somewhere in the mid 90s.
Where exactly does CR get its statistics from? "Thin air" seems a
possible answer.
JS
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
Christopher Wong wrote:
> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>> Christopher Wong wrote:
>>> The reason for the reliability disparity is quite simple. While Hyundai has
>>> made remarkable progress in reliability, they may not be as consistent
>>> across all models or when first starting out. The Tucson is still very new,
>> For Hyundai, IMHO, I think you have it backwards. They are making
>> strides in quality very quickly, and the trend seems to be the newer
>> the design the better the build. This does not surprise me, the Koreans
>> are world leaders in computerized design, manufacturing and automation.
>> This may be why the newer models are doing so well in JD Power (fresher
>> data) and not Consumer Reports (last years data). In fact, CR wrote
>> that the predicted reliability of the 2006 Tucson will be "worse than
>> average." From what we know today, that prediction has proven to be
>> dead wrong.
>
> I say you are way too optimistic -- you are predicting the future -- based
> on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
> can't extrapolate reliability from this study. Nor should it suggest that
> the current Tucson should improve significantly: why do you assume it
> improved its IQS score? It has always done well. According to Edmunds, the
> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
> on the IQS. So naively going by the unchanged IQS score, we can assume that
> the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
> the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
> category, tying with the Kia Sportage and Saturn Vue.
Consumer Reports, the people that put the Toyota Corolla and the Geo
Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
the early 90s.
> While I am optimistic about Hyundai's quality in general, it sure looks like
> the Tucson is a dud. This is the first Hyundai in years that scored this
> horribly. The current Sonata debuted on the charts with an above-average
> rating. The Santa Fe was always respectable across all years. The old Accent
> and Elantra never scored this bad out the door. My fear is that there is
> something badly wrong with the Tucson. You seem to be expecting miracles to
> happen with the Tucson, but Hyundai simply does not have that kind of track
> record. The Sonata is reliable now, but when it was introduced in 2000 it
> spent the first 2 model years with below-average reliability. There is
> absolutely no proof that Hyundai turned around the Tucson in just one year.
The Sonata was introduced long before 2000... I want to say 1989 or
1990. The Sonata V6 was introduced somewhere in the mid 90s.
Where exactly does CR get its statistics from? "Thin air" seems a
possible answer.
JS
> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>> Christopher Wong wrote:
>>> The reason for the reliability disparity is quite simple. While Hyundai has
>>> made remarkable progress in reliability, they may not be as consistent
>>> across all models or when first starting out. The Tucson is still very new,
>> For Hyundai, IMHO, I think you have it backwards. They are making
>> strides in quality very quickly, and the trend seems to be the newer
>> the design the better the build. This does not surprise me, the Koreans
>> are world leaders in computerized design, manufacturing and automation.
>> This may be why the newer models are doing so well in JD Power (fresher
>> data) and not Consumer Reports (last years data). In fact, CR wrote
>> that the predicted reliability of the 2006 Tucson will be "worse than
>> average." From what we know today, that prediction has proven to be
>> dead wrong.
>
> I say you are way too optimistic -- you are predicting the future -- based
> on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
> can't extrapolate reliability from this study. Nor should it suggest that
> the current Tucson should improve significantly: why do you assume it
> improved its IQS score? It has always done well. According to Edmunds, the
> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
> on the IQS. So naively going by the unchanged IQS score, we can assume that
> the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
> the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
> category, tying with the Kia Sportage and Saturn Vue.
Consumer Reports, the people that put the Toyota Corolla and the Geo
Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
the early 90s.
> While I am optimistic about Hyundai's quality in general, it sure looks like
> the Tucson is a dud. This is the first Hyundai in years that scored this
> horribly. The current Sonata debuted on the charts with an above-average
> rating. The Santa Fe was always respectable across all years. The old Accent
> and Elantra never scored this bad out the door. My fear is that there is
> something badly wrong with the Tucson. You seem to be expecting miracles to
> happen with the Tucson, but Hyundai simply does not have that kind of track
> record. The Sonata is reliable now, but when it was introduced in 2000 it
> spent the first 2 model years with below-average reliability. There is
> absolutely no proof that Hyundai turned around the Tucson in just one year.
The Sonata was introduced long before 2000... I want to say 1989 or
1990. The Sonata V6 was introduced somewhere in the mid 90s.
Where exactly does CR get its statistics from? "Thin air" seems a
possible answer.
JS
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
Christopher Wong wrote:
> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>> Christopher Wong wrote:
>>> The reason for the reliability disparity is quite simple. While Hyundai has
>>> made remarkable progress in reliability, they may not be as consistent
>>> across all models or when first starting out. The Tucson is still very new,
>> For Hyundai, IMHO, I think you have it backwards. They are making
>> strides in quality very quickly, and the trend seems to be the newer
>> the design the better the build. This does not surprise me, the Koreans
>> are world leaders in computerized design, manufacturing and automation.
>> This may be why the newer models are doing so well in JD Power (fresher
>> data) and not Consumer Reports (last years data). In fact, CR wrote
>> that the predicted reliability of the 2006 Tucson will be "worse than
>> average." From what we know today, that prediction has proven to be
>> dead wrong.
>
> I say you are way too optimistic -- you are predicting the future -- based
> on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
> can't extrapolate reliability from this study. Nor should it suggest that
> the current Tucson should improve significantly: why do you assume it
> improved its IQS score? It has always done well. According to Edmunds, the
> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
> on the IQS. So naively going by the unchanged IQS score, we can assume that
> the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
> the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
> category, tying with the Kia Sportage and Saturn Vue.
Consumer Reports, the people that put the Toyota Corolla and the Geo
Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
the early 90s.
> While I am optimistic about Hyundai's quality in general, it sure looks like
> the Tucson is a dud. This is the first Hyundai in years that scored this
> horribly. The current Sonata debuted on the charts with an above-average
> rating. The Santa Fe was always respectable across all years. The old Accent
> and Elantra never scored this bad out the door. My fear is that there is
> something badly wrong with the Tucson. You seem to be expecting miracles to
> happen with the Tucson, but Hyundai simply does not have that kind of track
> record. The Sonata is reliable now, but when it was introduced in 2000 it
> spent the first 2 model years with below-average reliability. There is
> absolutely no proof that Hyundai turned around the Tucson in just one year.
The Sonata was introduced long before 2000... I want to say 1989 or
1990. The Sonata V6 was introduced somewhere in the mid 90s.
Where exactly does CR get its statistics from? "Thin air" seems a
possible answer.
JS
> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>> Christopher Wong wrote:
>>> The reason for the reliability disparity is quite simple. While Hyundai has
>>> made remarkable progress in reliability, they may not be as consistent
>>> across all models or when first starting out. The Tucson is still very new,
>> For Hyundai, IMHO, I think you have it backwards. They are making
>> strides in quality very quickly, and the trend seems to be the newer
>> the design the better the build. This does not surprise me, the Koreans
>> are world leaders in computerized design, manufacturing and automation.
>> This may be why the newer models are doing so well in JD Power (fresher
>> data) and not Consumer Reports (last years data). In fact, CR wrote
>> that the predicted reliability of the 2006 Tucson will be "worse than
>> average." From what we know today, that prediction has proven to be
>> dead wrong.
>
> I say you are way too optimistic -- you are predicting the future -- based
> on wrong data. As I said, the JD Power IQS is NOT a reliability survey. You
> can't extrapolate reliability from this study. Nor should it suggest that
> the current Tucson should improve significantly: why do you assume it
> improved its IQS score? It has always done well. According to Edmunds, the
> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores 4/5
> on the IQS. So naively going by the unchanged IQS score, we can assume that
> the 2006 Tucson should be about as reliable as the 2005. Unfortunately, in
> the CR reliability survey the 2005 Tucson ranks dead last in the small SUV
> category, tying with the Kia Sportage and Saturn Vue.
Consumer Reports, the people that put the Toyota Corolla and the Geo
Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
the early 90s.
> While I am optimistic about Hyundai's quality in general, it sure looks like
> the Tucson is a dud. This is the first Hyundai in years that scored this
> horribly. The current Sonata debuted on the charts with an above-average
> rating. The Santa Fe was always respectable across all years. The old Accent
> and Elantra never scored this bad out the door. My fear is that there is
> something badly wrong with the Tucson. You seem to be expecting miracles to
> happen with the Tucson, but Hyundai simply does not have that kind of track
> record. The Sonata is reliable now, but when it was introduced in 2000 it
> spent the first 2 model years with below-average reliability. There is
> absolutely no proof that Hyundai turned around the Tucson in just one year.
The Sonata was introduced long before 2000... I want to say 1989 or
1990. The Sonata V6 was introduced somewhere in the mid 90s.
Where exactly does CR get its statistics from? "Thin air" seems a
possible answer.
JS
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
JS wrote:
> Christopher Wong wrote:
>
>> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>>
>>> Christopher Wong wrote:
>>>
>>>> The reason for the reliability disparity is quite simple. While
>>>> Hyundai has
>>>> made remarkable progress in reliability, they may not be as consistent
>>>> across all models or when first starting out. The Tucson is still
>>>> very new,
>>>
>>> For Hyundai, IMHO, I think you have it backwards. They are making
>>> strides in quality very quickly, and the trend seems to be the newer
>>> the design the better the build. This does not surprise me, the Koreans
>>> are world leaders in computerized design, manufacturing and automation.
>>> This may be why the newer models are doing so well in JD Power (fresher
>>> data) and not Consumer Reports (last years data). In fact, CR wrote
>>> that the predicted reliability of the 2006 Tucson will be "worse than
>>> average." From what we know today, that prediction has proven to be
>>> dead wrong.
>>
>>
>> I say you are way too optimistic -- you are predicting the future --
>> based
>> on wrong data. As I said, the JD Power IQS is NOT a reliability
>> survey. You
>> can't extrapolate reliability from this study. Nor should it suggest that
>> the current Tucson should improve significantly: why do you assume it
>> improved its IQS score? It has always done well. According to Edmunds,
>> the
>> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores
>> 4/5
>> on the IQS. So naively going by the unchanged IQS score, we can assume
>> that
>> the 2006 Tucson should be about as reliable as the 2005.
>> Unfortunately, in
>> the CR reliability survey the 2005 Tucson ranks dead last in the small
>> SUV
>> category, tying with the Kia Sportage and Saturn Vue.
>
>
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
Yes, Consumer Reports data isn't from a randomly selected sample, it is
from a self-selected set of their subscribers. So you have at least two
biases built in to start with. The bias of having only those people who
subscriber and then the bias of having only those who choose to return
the survey.
I've owned two cars that were also rated pretty much at the opposite end
of CR's spectrum: an 84 Accord that rated well and an 89 Acclaim that
rated poorly. The Accord was a lemon (engine failed at less than 80,000
miles - only engine I've had in 30 years of driving that failed) and the
Acclaim was the rock of Gibralter.
Matt
> Christopher Wong wrote:
>
>> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>>
>>> Christopher Wong wrote:
>>>
>>>> The reason for the reliability disparity is quite simple. While
>>>> Hyundai has
>>>> made remarkable progress in reliability, they may not be as consistent
>>>> across all models or when first starting out. The Tucson is still
>>>> very new,
>>>
>>> For Hyundai, IMHO, I think you have it backwards. They are making
>>> strides in quality very quickly, and the trend seems to be the newer
>>> the design the better the build. This does not surprise me, the Koreans
>>> are world leaders in computerized design, manufacturing and automation.
>>> This may be why the newer models are doing so well in JD Power (fresher
>>> data) and not Consumer Reports (last years data). In fact, CR wrote
>>> that the predicted reliability of the 2006 Tucson will be "worse than
>>> average." From what we know today, that prediction has proven to be
>>> dead wrong.
>>
>>
>> I say you are way too optimistic -- you are predicting the future --
>> based
>> on wrong data. As I said, the JD Power IQS is NOT a reliability
>> survey. You
>> can't extrapolate reliability from this study. Nor should it suggest that
>> the current Tucson should improve significantly: why do you assume it
>> improved its IQS score? It has always done well. According to Edmunds,
>> the
>> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores
>> 4/5
>> on the IQS. So naively going by the unchanged IQS score, we can assume
>> that
>> the 2006 Tucson should be about as reliable as the 2005.
>> Unfortunately, in
>> the CR reliability survey the 2005 Tucson ranks dead last in the small
>> SUV
>> category, tying with the Kia Sportage and Saturn Vue.
>
>
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
Yes, Consumer Reports data isn't from a randomly selected sample, it is
from a self-selected set of their subscribers. So you have at least two
biases built in to start with. The bias of having only those people who
subscriber and then the bias of having only those who choose to return
the survey.
I've owned two cars that were also rated pretty much at the opposite end
of CR's spectrum: an 84 Accord that rated well and an 89 Acclaim that
rated poorly. The Accord was a lemon (engine failed at less than 80,000
miles - only engine I've had in 30 years of driving that failed) and the
Acclaim was the rock of Gibralter.
Matt
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
JS wrote:
> Christopher Wong wrote:
>
>> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>>
>>> Christopher Wong wrote:
>>>
>>>> The reason for the reliability disparity is quite simple. While
>>>> Hyundai has
>>>> made remarkable progress in reliability, they may not be as consistent
>>>> across all models or when first starting out. The Tucson is still
>>>> very new,
>>>
>>> For Hyundai, IMHO, I think you have it backwards. They are making
>>> strides in quality very quickly, and the trend seems to be the newer
>>> the design the better the build. This does not surprise me, the Koreans
>>> are world leaders in computerized design, manufacturing and automation.
>>> This may be why the newer models are doing so well in JD Power (fresher
>>> data) and not Consumer Reports (last years data). In fact, CR wrote
>>> that the predicted reliability of the 2006 Tucson will be "worse than
>>> average." From what we know today, that prediction has proven to be
>>> dead wrong.
>>
>>
>> I say you are way too optimistic -- you are predicting the future --
>> based
>> on wrong data. As I said, the JD Power IQS is NOT a reliability
>> survey. You
>> can't extrapolate reliability from this study. Nor should it suggest that
>> the current Tucson should improve significantly: why do you assume it
>> improved its IQS score? It has always done well. According to Edmunds,
>> the
>> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores
>> 4/5
>> on the IQS. So naively going by the unchanged IQS score, we can assume
>> that
>> the 2006 Tucson should be about as reliable as the 2005.
>> Unfortunately, in
>> the CR reliability survey the 2005 Tucson ranks dead last in the small
>> SUV
>> category, tying with the Kia Sportage and Saturn Vue.
>
>
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
Yes, Consumer Reports data isn't from a randomly selected sample, it is
from a self-selected set of their subscribers. So you have at least two
biases built in to start with. The bias of having only those people who
subscriber and then the bias of having only those who choose to return
the survey.
I've owned two cars that were also rated pretty much at the opposite end
of CR's spectrum: an 84 Accord that rated well and an 89 Acclaim that
rated poorly. The Accord was a lemon (engine failed at less than 80,000
miles - only engine I've had in 30 years of driving that failed) and the
Acclaim was the rock of Gibralter.
Matt
> Christopher Wong wrote:
>
>> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>>
>>> Christopher Wong wrote:
>>>
>>>> The reason for the reliability disparity is quite simple. While
>>>> Hyundai has
>>>> made remarkable progress in reliability, they may not be as consistent
>>>> across all models or when first starting out. The Tucson is still
>>>> very new,
>>>
>>> For Hyundai, IMHO, I think you have it backwards. They are making
>>> strides in quality very quickly, and the trend seems to be the newer
>>> the design the better the build. This does not surprise me, the Koreans
>>> are world leaders in computerized design, manufacturing and automation.
>>> This may be why the newer models are doing so well in JD Power (fresher
>>> data) and not Consumer Reports (last years data). In fact, CR wrote
>>> that the predicted reliability of the 2006 Tucson will be "worse than
>>> average." From what we know today, that prediction has proven to be
>>> dead wrong.
>>
>>
>> I say you are way too optimistic -- you are predicting the future --
>> based
>> on wrong data. As I said, the JD Power IQS is NOT a reliability
>> survey. You
>> can't extrapolate reliability from this study. Nor should it suggest that
>> the current Tucson should improve significantly: why do you assume it
>> improved its IQS score? It has always done well. According to Edmunds,
>> the
>> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores
>> 4/5
>> on the IQS. So naively going by the unchanged IQS score, we can assume
>> that
>> the 2006 Tucson should be about as reliable as the 2005.
>> Unfortunately, in
>> the CR reliability survey the 2005 Tucson ranks dead last in the small
>> SUV
>> category, tying with the Kia Sportage and Saturn Vue.
>
>
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
Yes, Consumer Reports data isn't from a randomly selected sample, it is
from a self-selected set of their subscribers. So you have at least two
biases built in to start with. The bias of having only those people who
subscriber and then the bias of having only those who choose to return
the survey.
I've owned two cars that were also rated pretty much at the opposite end
of CR's spectrum: an 84 Accord that rated well and an 89 Acclaim that
rated poorly. The Accord was a lemon (engine failed at less than 80,000
miles - only engine I've had in 30 years of driving that failed) and the
Acclaim was the rock of Gibralter.
Matt
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
JS wrote:
> Christopher Wong wrote:
>
>> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>>
>>> Christopher Wong wrote:
>>>
>>>> The reason for the reliability disparity is quite simple. While
>>>> Hyundai has
>>>> made remarkable progress in reliability, they may not be as consistent
>>>> across all models or when first starting out. The Tucson is still
>>>> very new,
>>>
>>> For Hyundai, IMHO, I think you have it backwards. They are making
>>> strides in quality very quickly, and the trend seems to be the newer
>>> the design the better the build. This does not surprise me, the Koreans
>>> are world leaders in computerized design, manufacturing and automation.
>>> This may be why the newer models are doing so well in JD Power (fresher
>>> data) and not Consumer Reports (last years data). In fact, CR wrote
>>> that the predicted reliability of the 2006 Tucson will be "worse than
>>> average." From what we know today, that prediction has proven to be
>>> dead wrong.
>>
>>
>> I say you are way too optimistic -- you are predicting the future --
>> based
>> on wrong data. As I said, the JD Power IQS is NOT a reliability
>> survey. You
>> can't extrapolate reliability from this study. Nor should it suggest that
>> the current Tucson should improve significantly: why do you assume it
>> improved its IQS score? It has always done well. According to Edmunds,
>> the
>> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores
>> 4/5
>> on the IQS. So naively going by the unchanged IQS score, we can assume
>> that
>> the 2006 Tucson should be about as reliable as the 2005.
>> Unfortunately, in
>> the CR reliability survey the 2005 Tucson ranks dead last in the small
>> SUV
>> category, tying with the Kia Sportage and Saturn Vue.
>
>
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
Yes, Consumer Reports data isn't from a randomly selected sample, it is
from a self-selected set of their subscribers. So you have at least two
biases built in to start with. The bias of having only those people who
subscriber and then the bias of having only those who choose to return
the survey.
I've owned two cars that were also rated pretty much at the opposite end
of CR's spectrum: an 84 Accord that rated well and an 89 Acclaim that
rated poorly. The Accord was a lemon (engine failed at less than 80,000
miles - only engine I've had in 30 years of driving that failed) and the
Acclaim was the rock of Gibralter.
Matt
> Christopher Wong wrote:
>
>> On 2006-07-30, James <wimpyVO2max@gmail.com> wrote:
>>
>>> Christopher Wong wrote:
>>>
>>>> The reason for the reliability disparity is quite simple. While
>>>> Hyundai has
>>>> made remarkable progress in reliability, they may not be as consistent
>>>> across all models or when first starting out. The Tucson is still
>>>> very new,
>>>
>>> For Hyundai, IMHO, I think you have it backwards. They are making
>>> strides in quality very quickly, and the trend seems to be the newer
>>> the design the better the build. This does not surprise me, the Koreans
>>> are world leaders in computerized design, manufacturing and automation.
>>> This may be why the newer models are doing so well in JD Power (fresher
>>> data) and not Consumer Reports (last years data). In fact, CR wrote
>>> that the predicted reliability of the 2006 Tucson will be "worse than
>>> average." From what we know today, that prediction has proven to be
>>> dead wrong.
>>
>>
>> I say you are way too optimistic -- you are predicting the future --
>> based
>> on wrong data. As I said, the JD Power IQS is NOT a reliability
>> survey. You
>> can't extrapolate reliability from this study. Nor should it suggest that
>> the current Tucson should improve significantly: why do you assume it
>> improved its IQS score? It has always done well. According to Edmunds,
>> the
>> 2005 Tucson scored 4/5 overall on the IQS. The 2006 Tucson also scores
>> 4/5
>> on the IQS. So naively going by the unchanged IQS score, we can assume
>> that
>> the 2006 Tucson should be about as reliable as the 2005.
>> Unfortunately, in
>> the CR reliability survey the 2005 Tucson ranks dead last in the small
>> SUV
>> category, tying with the Kia Sportage and Saturn Vue.
>
>
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
Yes, Consumer Reports data isn't from a randomly selected sample, it is
from a self-selected set of their subscribers. So you have at least two
biases built in to start with. The bias of having only those people who
subscriber and then the bias of having only those who choose to return
the survey.
I've owned two cars that were also rated pretty much at the opposite end
of CR's spectrum: an 84 Accord that rated well and an 89 Acclaim that
rated poorly. The Accord was a lemon (engine failed at less than 80,000
miles - only engine I've had in 30 years of driving that failed) and the
Acclaim was the rock of Gibralter.
Matt
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
On 2006-08-02, Matt Whiting <whiting@epix.net> wrote:
> Yes, Consumer Reports data isn't from a randomly selected sample, it is
> from a self-selected set of their subscribers. So you have at least two
> biases built in to start with. The bias of having only those people who
> subscriber and then the bias of having only those who choose to return
> the survey.
You realize, you have defined away the entire field of surveying. There
will always be some selection issues in statistical sampling, precisely
because it is sampling. There will always be the bias of having getting only
results from people willing to return results. You won't get that
theoretical purity without testing absolutely everyone and forcing them to
submit an answer. Then that is no longer a survey.
Valid sampling simply means to make sure that the sampling does not
significantly distort the results. And here, the CR bashers' arguments
evaporate. Nobody has been able to prove that -- or even explain how -- CR's
sampling distorts in its stated purpose of comparative reliablity. I repeat:
comparative reliability. Why should a CR reader who owns a Tucson be more
biased against his car than a CR reader who owns a Santa Fe? Feh.
CR's survey has a sample size of about a million: 20 times the size of JD
Powers' 3 year reliability survey. Its results are empirical: it counts the
problem incident rate. I don't see how there an be a statistical or bias
fluke when the Tucson and Santa Fe's ratings are so far apart. I believe the
CR's sample size has grown to the point where its numbers are fairly
consistent for the more popular models. Case in point: the Tucson's
reliability score is virtually identical to the Kia Sportage, its mechanical
twin.
Chris
> Yes, Consumer Reports data isn't from a randomly selected sample, it is
> from a self-selected set of their subscribers. So you have at least two
> biases built in to start with. The bias of having only those people who
> subscriber and then the bias of having only those who choose to return
> the survey.
You realize, you have defined away the entire field of surveying. There
will always be some selection issues in statistical sampling, precisely
because it is sampling. There will always be the bias of having getting only
results from people willing to return results. You won't get that
theoretical purity without testing absolutely everyone and forcing them to
submit an answer. Then that is no longer a survey.
Valid sampling simply means to make sure that the sampling does not
significantly distort the results. And here, the CR bashers' arguments
evaporate. Nobody has been able to prove that -- or even explain how -- CR's
sampling distorts in its stated purpose of comparative reliablity. I repeat:
comparative reliability. Why should a CR reader who owns a Tucson be more
biased against his car than a CR reader who owns a Santa Fe? Feh.
CR's survey has a sample size of about a million: 20 times the size of JD
Powers' 3 year reliability survey. Its results are empirical: it counts the
problem incident rate. I don't see how there an be a statistical or bias
fluke when the Tucson and Santa Fe's ratings are so far apart. I believe the
CR's sample size has grown to the point where its numbers are fairly
consistent for the more popular models. Case in point: the Tucson's
reliability score is virtually identical to the Kia Sportage, its mechanical
twin.
Chris
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
On 2006-08-02, Matt Whiting <whiting@epix.net> wrote:
> Yes, Consumer Reports data isn't from a randomly selected sample, it is
> from a self-selected set of their subscribers. So you have at least two
> biases built in to start with. The bias of having only those people who
> subscriber and then the bias of having only those who choose to return
> the survey.
You realize, you have defined away the entire field of surveying. There
will always be some selection issues in statistical sampling, precisely
because it is sampling. There will always be the bias of having getting only
results from people willing to return results. You won't get that
theoretical purity without testing absolutely everyone and forcing them to
submit an answer. Then that is no longer a survey.
Valid sampling simply means to make sure that the sampling does not
significantly distort the results. And here, the CR bashers' arguments
evaporate. Nobody has been able to prove that -- or even explain how -- CR's
sampling distorts in its stated purpose of comparative reliablity. I repeat:
comparative reliability. Why should a CR reader who owns a Tucson be more
biased against his car than a CR reader who owns a Santa Fe? Feh.
CR's survey has a sample size of about a million: 20 times the size of JD
Powers' 3 year reliability survey. Its results are empirical: it counts the
problem incident rate. I don't see how there an be a statistical or bias
fluke when the Tucson and Santa Fe's ratings are so far apart. I believe the
CR's sample size has grown to the point where its numbers are fairly
consistent for the more popular models. Case in point: the Tucson's
reliability score is virtually identical to the Kia Sportage, its mechanical
twin.
Chris
> Yes, Consumer Reports data isn't from a randomly selected sample, it is
> from a self-selected set of their subscribers. So you have at least two
> biases built in to start with. The bias of having only those people who
> subscriber and then the bias of having only those who choose to return
> the survey.
You realize, you have defined away the entire field of surveying. There
will always be some selection issues in statistical sampling, precisely
because it is sampling. There will always be the bias of having getting only
results from people willing to return results. You won't get that
theoretical purity without testing absolutely everyone and forcing them to
submit an answer. Then that is no longer a survey.
Valid sampling simply means to make sure that the sampling does not
significantly distort the results. And here, the CR bashers' arguments
evaporate. Nobody has been able to prove that -- or even explain how -- CR's
sampling distorts in its stated purpose of comparative reliablity. I repeat:
comparative reliability. Why should a CR reader who owns a Tucson be more
biased against his car than a CR reader who owns a Santa Fe? Feh.
CR's survey has a sample size of about a million: 20 times the size of JD
Powers' 3 year reliability survey. Its results are empirical: it counts the
problem incident rate. I don't see how there an be a statistical or bias
fluke when the Tucson and Santa Fe's ratings are so far apart. I believe the
CR's sample size has grown to the point where its numbers are fairly
consistent for the more popular models. Case in point: the Tucson's
reliability score is virtually identical to the Kia Sportage, its mechanical
twin.
Chris
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
On 2006-08-02, Matt Whiting <whiting@epix.net> wrote:
> Yes, Consumer Reports data isn't from a randomly selected sample, it is
> from a self-selected set of their subscribers. So you have at least two
> biases built in to start with. The bias of having only those people who
> subscriber and then the bias of having only those who choose to return
> the survey.
You realize, you have defined away the entire field of surveying. There
will always be some selection issues in statistical sampling, precisely
because it is sampling. There will always be the bias of having getting only
results from people willing to return results. You won't get that
theoretical purity without testing absolutely everyone and forcing them to
submit an answer. Then that is no longer a survey.
Valid sampling simply means to make sure that the sampling does not
significantly distort the results. And here, the CR bashers' arguments
evaporate. Nobody has been able to prove that -- or even explain how -- CR's
sampling distorts in its stated purpose of comparative reliablity. I repeat:
comparative reliability. Why should a CR reader who owns a Tucson be more
biased against his car than a CR reader who owns a Santa Fe? Feh.
CR's survey has a sample size of about a million: 20 times the size of JD
Powers' 3 year reliability survey. Its results are empirical: it counts the
problem incident rate. I don't see how there an be a statistical or bias
fluke when the Tucson and Santa Fe's ratings are so far apart. I believe the
CR's sample size has grown to the point where its numbers are fairly
consistent for the more popular models. Case in point: the Tucson's
reliability score is virtually identical to the Kia Sportage, its mechanical
twin.
Chris
> Yes, Consumer Reports data isn't from a randomly selected sample, it is
> from a self-selected set of their subscribers. So you have at least two
> biases built in to start with. The bias of having only those people who
> subscriber and then the bias of having only those who choose to return
> the survey.
You realize, you have defined away the entire field of surveying. There
will always be some selection issues in statistical sampling, precisely
because it is sampling. There will always be the bias of having getting only
results from people willing to return results. You won't get that
theoretical purity without testing absolutely everyone and forcing them to
submit an answer. Then that is no longer a survey.
Valid sampling simply means to make sure that the sampling does not
significantly distort the results. And here, the CR bashers' arguments
evaporate. Nobody has been able to prove that -- or even explain how -- CR's
sampling distorts in its stated purpose of comparative reliablity. I repeat:
comparative reliability. Why should a CR reader who owns a Tucson be more
biased against his car than a CR reader who owns a Santa Fe? Feh.
CR's survey has a sample size of about a million: 20 times the size of JD
Powers' 3 year reliability survey. Its results are empirical: it counts the
problem incident rate. I don't see how there an be a statistical or bias
fluke when the Tucson and Santa Fe's ratings are so far apart. I believe the
CR's sample size has grown to the point where its numbers are fairly
consistent for the more popular models. Case in point: the Tucson's
reliability score is virtually identical to the Kia Sportage, its mechanical
twin.
Chris
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
JS wrote:
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
>
CR doesn't rate the reliability, they report the statistics based on
what the readers report. If the Toyota scored higher, then it was
because the Toyota owners reported fewer problems than average in the
survey.
>
> Where exactly does CR get its statistics from? "Thin air" seems a
> possible answer.
>
Statistics come from surveys sent out to CR subscribers. CR just
compiles the data and reports what the surveys say.
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
>
CR doesn't rate the reliability, they report the statistics based on
what the readers report. If the Toyota scored higher, then it was
because the Toyota owners reported fewer problems than average in the
survey.
>
> Where exactly does CR get its statistics from? "Thin air" seems a
> possible answer.
>
Statistics come from surveys sent out to CR subscribers. CR just
compiles the data and reports what the surveys say.
Guest
Posts: n/a
Re: Is the Tucson all that much better than the Santa Fe ?
JS wrote:
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
>
CR doesn't rate the reliability, they report the statistics based on
what the readers report. If the Toyota scored higher, then it was
because the Toyota owners reported fewer problems than average in the
survey.
>
> Where exactly does CR get its statistics from? "Thin air" seems a
> possible answer.
>
Statistics come from surveys sent out to CR subscribers. CR just
compiles the data and reports what the surveys say.
> Consumer Reports, the people that put the Toyota Corolla and the Geo
> Prism on opposite ends of the reliability scale (EXACT SAME VEHICLE) in
> the early 90s.
>
CR doesn't rate the reliability, they report the statistics based on
what the readers report. If the Toyota scored higher, then it was
because the Toyota owners reported fewer problems than average in the
survey.
>
> Where exactly does CR get its statistics from? "Thin air" seems a
> possible answer.
>
Statistics come from surveys sent out to CR subscribers. CR just
compiles the data and reports what the surveys say.
