Fuel Consumption Survey [Academic Survey]

• 

  brendanm on 14/09/2020 - 22:42

  -2

  Manufacturer fuel consumption is based on perfect lab conditions, from memory on a dyno. It doesn't factor in wind resistance or a million other variables. It is only useful for comparing different cars, not to get a real world consumption figure.

  Feel free to PayPal me money for doing your research for you.

  • 

    SniperFett on 14/09/2020 - 22:52

    Sorry for not clarifying in post, I am finding new methods such as real road testing to compare against manufacturer values and developing a procedure which would more accurately represent real world figures.

    • 

      resisting the urge on 14/09/2020 - 23:10

      You want to crowd source the work, but not the methods.

      And do it online. LOL.

      Bots are even easier to hire than collecting your data, and the form results can be whatever you want them to be. It only takes one person to decide they don't like you or your survey, and the whole thing will be as worthless as the way you started it…

    • 

      brad1-8tsi on 15/09/2020 - 00:03

      +1

      They do real road testing against a control tyre to give tyres a wear rating. It's very inaccurate.

      Even if you came up with the perfect test it would still be inaccurate due to real world variables. Don't even think about using the fuel light or DTE - they are on crack.

    • 

      brendanm on 15/09/2020 - 06:37

      +4

      www.fuelly.com

      Real road testing isn't a new method. It is extremely.variable though. Even the same car, same tyres etc, doing the same route, will have different consumption with different drivers. That's why they do the testing under lab conditions. The results aren't accurate compares to real.life consumption, but it does allow you to compare between different vehicles.

• 

  JIMB0 on 15/09/2020 - 05:44

  Search car forums. You'll find endless threads where people post up their fuel consumption figures. They vary widely even for the exact same car/engine/trans.

  • 

    Euphemistic on 15/09/2020 - 07:56

    +1

    They vary so wildly because the driver is the biggest contributor to fuel consumption variation. The way some lunatics race off from the lights and brake heavily at the next set it’s no wonder people can’t get the rated consumption. Then you add passengers, accessories, maintenance and a whole host of other factors.

    At least lab testing means it’s comparing apples and apples.

    For what it’s worth my vehicle gets near the rated consumption most of the time.

• 

  [Deactivated] on 15/09/2020 - 08:07

  +1

  There will never be a "real world" accurate figure, there are too many variables to make a generic figure.
  these include, but not limited to;
  tyre pressures
  wheel alignment
  load in vehicle
  tune of engine
  last oil change
  fuel type used
  topographical area driven
  speed driven
  wind speed and direction
  accelerator depression speed
  traffic conditions
  road raging
  and most importantly, the loose nut behind the wheel

  The government test is the most accurate for comparing models cause the computer controlling it doesn't get emotional when someone cuts them off

• 

  oscargamer on 15/09/2020 - 09:15

  As already said fuelly.com

• 

  toniyellow on 15/09/2020 - 10:03

  If this is a really academic study you need a participant information sheet and a lot more information that is available here, you can't draw any conclusions from a study that wasn't done properly.

  Your institution should have something like
  https://i.unisa.edu.au/staff/research/research-ethics/human-…
  to guide you.

  The survey may be short but there is a considerable amount of friction in getting people to monitor their fuel consumption even just once, so you should consider a prize draw or some other incentive. The research ethics unit at your institution should have suggestions on incentives.

• 

  pegaxs on 15/09/2020 - 10:46

  You are basically trying to reinvent the wheel. The problem you will face is that real world figures, as everyone else has already chimed in with, is that there is just too much variation. You will never get consistent results that are good enough to be reliable. And as soon as you do get these results, there will always be that person or group of people that say "I don't get anywhere near those figures"

  The problem is, you need to single out a make and model of vehicle and then gets thousands of owners input into the results to get any meaningful data from it. Posting a 5 question survey that doesn't nail down any details of the vehicle, where it is driven, how it is driven, what the weather is like, what time they fuelled up (etc etc…) means that of the 30 or so people that do reply, even if they are all 100% serious replies, is not going to give you any useful data. (A bit like "89% of people in Australia support… " but they only asked 30 people their views.)

  The reason these tests are done in a lab is because they can control everything and set a standard. That standard is the parameters that cars must work within. This gives a stable base line under known conditions to calculate an average type of result. Then this test is completed in the exact same fashion by everyone else that makes cars. This means that no car is given an advantage or is at a disadvantage due to "real world" variables. The idea of the tests is to level the playing field, so when you go to compare cars, you know that they were both tested in the same way, under the same conditions, using the same parameters.

  Basically, what would be real world driving results for you would be no where near what I am getting if I owned exactly the same vehicle.

  (Side note… WTF is it with all the negs in this thread?)

• 

  Moderatorscrimshaw on 15/09/2020 - 14:23

  Thread closed by Request