mileage database for HCHII bi-modal?
#1
mileage database for HCHII bi-modal?
I'm no statistician, so excuse me if I'm messed up here.
Looking at the cars that have many entries (HCHII at over 237, PriusII at 692 currently), the PriusII appears to have a standard deviation. But the HCHII looks almost bi-modal to me.
Anyone care to venture why this is? Is it that the HCHII only has 1/3 the data points, and does not have enough points to look "normal", like some of the other vehicles with smaller databases, or is 232 enough points to get a true representation. If so, what do you think is affecting it?
Looking at the cars that have many entries (HCHII at over 237, PriusII at 692 currently), the PriusII appears to have a standard deviation. But the HCHII looks almost bi-modal to me.
Anyone care to venture why this is? Is it that the HCHII only has 1/3 the data points, and does not have enough points to look "normal", like some of the other vehicles with smaller databases, or is 232 enough points to get a true representation. If so, what do you think is affecting it?
Last edited by toast64; 06-20-2006 at 05:38 PM. Reason: sp
#2
Re: mileage database for HCHII bi-modal?
You may want to consider the mileage of the cars that are in the population. I know with my HCH II, there was an appreciable difference when I crossed 3500 miles.
I would be an interesting comparison in the population sets in the miles of the vehicles. We may find a correlation in the two - miles vs. mpg. It could be either engine break in or learning how to drive it......
Just my 2 cents.
I would be an interesting comparison in the population sets in the miles of the vehicles. We may find a correlation in the two - miles vs. mpg. It could be either engine break in or learning how to drive it......
Just my 2 cents.
#3
Honda has more fanatics than Toyota
It doubt any Honda partisan will agree, but I think Honda-and thr HCH ll has more fanatics than Toyota and Prius. It is hard to read that from the numbers, but those 2 "bumps" you noticed might be a hint of just that.
A Honda buyer is more likely to be a committed mpg fan than a Prius buyer, therefore you are going to have a "bump' at above average-the fanatics and near fanatics- and a bump below average-the non fanatics who bought it because it is a reliable vehicle that gets good mpg, but don't really see it as a competition.
Toyota Hybrid -Prius-buyers just aren't as committed as Honda buyers. No one "loves" Toyota in the same way folks "love" Honda.Someday,I will make an effort to prove this, but not today.
Luck,Charlie
A Honda buyer is more likely to be a committed mpg fan than a Prius buyer, therefore you are going to have a "bump' at above average-the fanatics and near fanatics- and a bump below average-the non fanatics who bought it because it is a reliable vehicle that gets good mpg, but don't really see it as a competition.
Toyota Hybrid -Prius-buyers just aren't as committed as Honda buyers. No one "loves" Toyota in the same way folks "love" Honda.Someday,I will make an effort to prove this, but not today.
Luck,Charlie
#4
Re: mileage database for HCHII bi-modal?
Originally Posted by toast64
I'm no statistician, so excuse me if I'm messed up here.
Looking at the cars that have many entries (HCHII at over 237, PriusII at 692 currently), the PriusII appears to have a standard deviation. But the HCHII looks almost bi-modal to me.
Anyone care to venture why this is? Is it that the HCHII only has 1/3 the data points, and does not have enough points to look "normal", like some of the other vehicles with smaller databases, or is 232 enough points to get a true representation. If so, what do you think is affecting it?
Looking at the cars that have many entries (HCHII at over 237, PriusII at 692 currently), the PriusII appears to have a standard deviation. But the HCHII looks almost bi-modal to me.
Anyone care to venture why this is? Is it that the HCHII only has 1/3 the data points, and does not have enough points to look "normal", like some of the other vehicles with smaller databases, or is 232 enough points to get a true representation. If so, what do you think is affecting it?
This will eliminate the likly out liers and give you a chance to concentrate on those who are making regular contributions. I would expect their data entries to be more accurate.
If the bi-modal curve persists, start looking for a variable that may be associated with one or the other humps. You may be on the verge of finding some previously undocumented phenomina.
GOOD LUCK!
Bob Wilson
Last edited by bwilson4web; 06-21-2006 at 09:46 AM.
#5
Re: Honda has more fanatics than Toyota
Originally Posted by phoebeisis
No one "loves" Toyota in the same way folks "love" Honda.
#6
Re: mileage database for HCHII bi-modal?
VERY quick glance impressions:
1) lots and lots of low-tank contributors starting at 0 miles. Initial tanks, showing the car off to friends, experimenting, etc. Those under 4-5 tanks or 3-4K miles need to be tossed out of this analysis.
2) The bins are in 3mpg segments- that's a wide range. The fact that one covers to just over the hwy estimate, and the other is 15% under it may be explained by this:
- The upper peak may be representative of folks who bought the car specifically for their commute/highway driving mpg, and are achieving just that. I would fall into that category- 30 out of 33 miles of my commute are open highway.
- The lower peak is possibly where the 'true' average driver is- using the car as a family car, and mixing in a lot of errands/city driving?
The TCH data is worth keeping an eye on- it might start to exhibit the same behavior as its set grows?
1) lots and lots of low-tank contributors starting at 0 miles. Initial tanks, showing the car off to friends, experimenting, etc. Those under 4-5 tanks or 3-4K miles need to be tossed out of this analysis.
2) The bins are in 3mpg segments- that's a wide range. The fact that one covers to just over the hwy estimate, and the other is 15% under it may be explained by this:
- The upper peak may be representative of folks who bought the car specifically for their commute/highway driving mpg, and are achieving just that. I would fall into that category- 30 out of 33 miles of my commute are open highway.
- The lower peak is possibly where the 'true' average driver is- using the car as a family car, and mixing in a lot of errands/city driving?
The TCH data is worth keeping an eye on- it might start to exhibit the same behavior as its set grows?
#7
Re: mileage database for HCHII bi-modal?
Here's a closer look at the FE of the HCHII.
Also, here is a link that shows the number of tanks at the various FEs.
https://www.greenhybrid.com/share/fi...HIIFEtanks.JPG
Last edited by Mr. Kite; 06-21-2006 at 11:32 AM.
#8
Re: mileage database for HCHII bi-modal?
Originally Posted by bwilson4web
... Have you considered importing the data to Excel and only looking at the data of those with 4 or more tank entries? ...
I've lost track of the thread that discussed how to download to Excel. I seem to remember it from several months ago, but can't find it. Can you point me to it?
Thanks.
#9
Re: mileage database for HCHII bi-modal?
Originally Posted by toast64
Bob,
I've lost track of the thread that discussed how to download to Excel. I seem to remember it from several months ago, but can't find it. Can you point me to it?
Thanks.
I've lost track of the thread that discussed how to download to Excel. I seem to remember it from several months ago, but can't find it. Can you point me to it?
Thanks.
#10
Re: mileage database for HCHII bi-modal?
Originally Posted by toast64
. . .I've lost track of the thread that discussed how to download to Excel. I seem to remember it from several months ago, but can't find it. Can you point me to it?
Thanks.
Thanks.
Bob Wilson