data entry error skewing prius' average?
#1
04-01-2007, 03:02 PM
data entry error skewing prius' average?
Does it bother anyone else that a prius user claims to have consumed 70,726 gallons of fuel to drive 348 miles? This is a single tank of gas, mind you.
https://www.greenhybrid.com/compare/.../car/2141.html
The median prius efficiency by looking at the overall graph appears to be around 49.5, however the calculated average is 47.5. I suspect this is because of this one erroneous tank entry, which essentially adds an overhead of 71k gallons of fuel. This is huge.
Does this bother anyone else, or am I the only one? Isn't there someone that can invalidate this entry? Also, since we know the fuel capacity of prius does not exceed 12 gallons, so shouldn't there be a cap on the entry form?
By the way, there's another bad entry, but not as eggregious. This one only (!) claims 75 gallons of fuel for a single tank.
https://www.greenhybrid.com/compare/.../car/4450.html
By the way, greetings to everyone. I got my new 2007 Seaside Pearl Prius 1.5 months ago and I love it.
Jim
https://www.greenhybrid.com/compare/.../car/2141.html
The median prius efficiency by looking at the overall graph appears to be around 49.5, however the calculated average is 47.5. I suspect this is because of this one erroneous tank entry, which essentially adds an overhead of 71k gallons of fuel. This is huge.
Does this bother anyone else, or am I the only one? Isn't there someone that can invalidate this entry? Also, since we know the fuel capacity of prius does not exceed 12 gallons, so shouldn't there be a cap on the entry form?
By the way, there's another bad entry, but not as eggregious. This one only (!) claims 75 gallons of fuel for a single tank.
https://www.greenhybrid.com/compare/.../car/4450.html
By the way, greetings to everyone. I got my new 2007 Seaside Pearl Prius 1.5 months ago and I love it.
Jim
#2
04-01-2007, 05:28 PM
Re: data entry error skewing prius' average?
Hi Jim,
Check the summary page and you'll notice only the middle 50% are used for the calculations:
https://www.greenhybrid.com/compare/mileage/
What I'd suggest is posting this or a similar note to the Web forum.
Bob Wilson
Does it bother anyone else that a prius user claims to have consumed 70,726 gallons of fuel to drive 348 miles? This is a single tank of gas, mind you.
https://www.greenhybrid.com/compare/.../car/2141.html
The median prius efficiency by looking at the overall graph appears to be around 49.5, however the calculated average is 47.5. I suspect this is because of this one erroneous tank entry, which essentially adds an overhead of 71k gallons of fuel. This is huge.
Does this bother anyone else, or am I the only one? . . .
https://www.greenhybrid.com/compare/.../car/2141.html
The median prius efficiency by looking at the overall graph appears to be around 49.5, however the calculated average is 47.5. I suspect this is because of this one erroneous tank entry, which essentially adds an overhead of 71k gallons of fuel. This is huge.
Does this bother anyone else, or am I the only one? . . .
https://www.greenhybrid.com/compare/mileage/
What I'd suggest is posting this or a similar note to the Web forum.
Bob Wilson
#3
04-01-2007, 10:47 PM
Re: data entry error skewing prius' average?
OK folks. I found that error. The last Prius, i.e. the one at the bottom of the ranking. Do you suppose someone could let him/her know? That entry really screws up their mpg.
Either take the avg of all the other entries, or delete that one. Of course, the best would be to correct it.
Too bad.
Either take the avg of all the other entries, or delete that one. Of course, the best would be to correct it.
Too bad.
#8
04-08-2007, 10:13 AM
Re: data entry error skewing prius' average?
I'm still a bit fuzzy on how these numbers are calculated. The summary page says you "should' use just the middle 50% to get an idea of what to expect.
Is the mean mpg the mean of each members tank? In which case the member with the error has a .1mpg average-essentially zero. If that is so, then he drops the average by about .05mpg..Assuming that the average of member's average tank is 47mpg, then to back calculate I did this
47mpg times 1140 members=53580 m/g xmembers
Then do the same with one more member contributing a 0mpg average
53580/1141= 46.95mpg.
If the average is actually an average of total miles traveled vs fuel used,I had to make some guesses.
Using 47mpg and assuming about 4700 miles per member, then each member would use about 100 gallons.
The total distance would be 4700 miles x1140= 5,358,000 miles and 114,000 gallons.
If we add the 70,000 gallons then it is 5,358,000/184,000= 29mpg.
This can't be the way it is calculated.
My guess is that the mean is calculated with each member contributing just one number-his average mpg calculated by dividing his total miles by his total fuel.In truth if we just used his tank average(add up his tank mpg and divide by his tanks) we would get about the same number.
In not so short, the goof up doesn't seem to matter too much.
I suspect it was a HCH2 member who hacked in(just kidding HCH2 members-the HCH2 is a fine car-second best, but nice).
Hmmm,I seem to have too much time on my hands.
Thanks,
Charlie
PS Bob,I have taken some crude pictures of your idea to use the front passenger's seat as sleep space; it will work great. I can do the same with the front drivers seat.THANKS!!
Is the mean mpg the mean of each members tank? In which case the member with the error has a .1mpg average-essentially zero. If that is so, then he drops the average by about .05mpg..Assuming that the average of member's average tank is 47mpg, then to back calculate I did this
47mpg times 1140 members=53580 m/g xmembers
Then do the same with one more member contributing a 0mpg average
53580/1141= 46.95mpg.
If the average is actually an average of total miles traveled vs fuel used,I had to make some guesses.
Using 47mpg and assuming about 4700 miles per member, then each member would use about 100 gallons.
The total distance would be 4700 miles x1140= 5,358,000 miles and 114,000 gallons.
If we add the 70,000 gallons then it is 5,358,000/184,000= 29mpg.
This can't be the way it is calculated.
My guess is that the mean is calculated with each member contributing just one number-his average mpg calculated by dividing his total miles by his total fuel.In truth if we just used his tank average(add up his tank mpg and divide by his tanks) we would get about the same number.
In not so short, the goof up doesn't seem to matter too much.
I suspect it was a HCH2 member who hacked in(just kidding HCH2 members-the HCH2 is a fine car-second best, but nice).
Hmmm,I seem to have too much time on my hands.
Thanks,
Charlie
PS Bob,I have taken some crude pictures of your idea to use the front passenger's seat as sleep space; it will work great. I can do the same with the front drivers seat.THANKS!!
#9
04-08-2007, 02:31 PM
Re: data entry error skewing prius' average?
I was curious, so I copied the summary page (the page with one record per car) into a spreadsheet and calculated the totals.
For each car, I calculated the gallons by dividing miles by miles per gallon.
Then I calculate the overall miles, and overall gallons, and divide the former by the latter to get the average mpg over all cars. The result is 36.19 mpg. If I exclude the last record, the average mpg rises to 48.2, which is a lot closer to the published average.
So it would seem that the bad record is not contributing to the overall average.
For each car, I calculated the gallons by dividing miles by miles per gallon.
Then I calculate the overall miles, and overall gallons, and divide the former by the latter to get the average mpg over all cars. The result is 36.19 mpg. If I exclude the last record, the average mpg rises to 48.2, which is a lot closer to the published average.
So it would seem that the bad record is not contributing to the overall average.
#10
04-08-2007, 02:39 PM
Re: data entry error skewing prius' average?
phoebeisis,
Apparently there are two ways to present the avg mpg (well, who knows, maybe more). One is using the display average, and the other is to calculate in the traditional manner. The database allows for both.
Everything else you've said is probably so. However, as a simple soul, I have the database calculate that average. I've found the display is not all that accurate.
Apparently there are two ways to present the avg mpg (well, who knows, maybe more). One is using the display average, and the other is to calculate in the traditional manner. The database allows for both.
Everything else you've said is probably so. However, as a simple soul, I have the database calculate that average. I've found the display is not all that accurate.
