Download:
pdf |
pdf ATTACHMENT A
FUEL ECONOMY INSTRUMENTATION STUDY
PHASE 1 AND PHASE 2
PROBLEM STATEMENT
Wide ranging claims are made for the effects of driver behavior on real-world fuel economy
under a variety of information and education conditions. These claims tend not to be backed by
accepted standards of research design—due in part no doubt to the difficulty and expense of
measuring real-world driving behaviors of any kind. In general claims to the efficacy of driver
prompts and training have not been based on comparison to control groups or carefully
constructed without/with research designs. Worse, these claims are often based on the results of
simulators in which the effects of learning the simulator are not separated from any effects due to
without/with conditions.
The purpose of the proposed research is to overcome as many of these difficulties as possible in
a straightforward test of the existence of real-world effects of fuel economy information
feedback to drivers. The research questions are as follows:
1. What is the distribution of real-world fuel economy increases that can be expected over
the long term from in-vehicle fuel economy displays (IVDs)? And if possible,
a. Do the distributions differ by vehicle type?
b. City/region?
2. How many people know driving style matters to fuel economy?
a. What do they think the range of effects is?
b. What do they think other consequences are?
3. If the desired effect of in-vehicle fuel economy displays IVDs is something that could be
called
“eco-driving,” then what proportion of the population are already eco-driving?
a. What is the effect of IVDs on eco-drivers? On non-eco-drivers?
STATEMENT OF WORK
To answer these questions a two-phase research program is proposed. This document briefly
describes both phases. A pilot and full study are linked through their effort to address the research
questions above. However, there are significant unknowns about the practical implementation of any
study to measure real-world driving behaviors. The first phase pilot study addresses many of these
unknowns to control the cost and duration of the second phase full study and to insure that the full
study does not suffer from the design shortcomings of prior work. A budget is presented for both the
pilot and full studies; the full study budget and final design though are outcomes of the pilot study.
The information presented here on the full study is for planning and scoping purposes only.
PHASE 1: PILOT STUDY
The research questions require a multi-method approach. First, measurement of the effects of
fuel economy information on real-world fuel economy requires a standardized display and a
means to record fuel economy. Second, whether drivers differ in predisposition, experience, or
learning regarding the research process can be ascertained in questionnaires and interviews. The
basic research process then is to engage households in a driving experiment in which a baseline
of fuel economy is established, driver feedback is provided and fuel economy measured again,
and finally, the household is interviewed regarding the experiment.
The pilot study will implement this basic design in nine households in Davis, CA. The purpose
of the pilot is to ascertain, formulate responses to, and test those responses to logistical,
technical, perceptual, and behavioral barriers to the research. Examples of potential problems to
be addressed include the fact that despite the display/recording devices to be deployed in the
research plug into the ODB port accessible in the passenger compartment in all passenger cars
and light-duty trucks sold in the U.S. since 1996, “access” is not uniformly implemented across
vehicles. The question is, can respondent households be expected to install, operate, and uninstall
the device themselves?
Additionally, the period of the without/with fuel economy information display conditions is
subject to conflicting pressures. Longer time periods allow for observing more fuel economy
data over more conditions and therefore the testing of smaller absolute differences as well as the
“averaging out” of more uncontrolled factors in the without/with time periods.1./ Conversely,
longer without/with periods drive up project cost and increase project timelines.
The pilot study will consist of the tasks listed below.
Equipment Procurement
o UC Davis will purchase and configure five (5) interface units capable of displaying
and recording fuel-economy related driving data and GPS signals.
Protocol Development
o UC Davis will develop a preliminary interface protocol for deployment of the units,
mail, phone, or in-person interviews, and an electronic survey that will be scalable for
the size of the final study given the assumed time resources of the study staff.
Sampling Protocol
o Volunteer participants will be recruited by AAA using a form letter sent to
qualifying AAA insurance policy holders in the Davis area as described below. In
addition, further screening will insure that participant vehicles match the study
criteria:
Vehicles do not currently display fuel-economy data
Vehicles are compatible with the study interface
Pilot Research
o UC Davis will deploy the interfaces in two to three waves to nine (9) total households
for a period of up to two months each. The purpose of this research will be to observe
measurable changes in driving behavior; to test the experimental protocol, and if
necessary modify the protocol between waves; to determine the minimum amount of
placement time required to observe measurable changes, with special attention paid to
the final goal of the full sample.
_______________________________________
1./ Longer time periods are not unambiguously better for averaging out uncontrolled factors. Letting a household drive
longer may simply give them the opportunity for that long vacation trip or some other difference to occur between the
without/with periods. One purpose of the post-interview is to explore for such differences with the households.
Fuel Economy Instrumentation Study
Attachment A
page 2 of 8
Data analysis and Report
o UC Davis will analyze the data from the nine pilot participants and prepare a report
with the pilot findings and experimental protocol recommendations for the full
sample. In particular, the report will specify:
Recommended full sample size
Recommended experimental protocol for the full sample including minimum
placement time
Recommendation to use or not use GPS-capable devices
Pilot fuel-economy results
Full sample study budget estimates
As noted above, the pilot study participants will be located in Davis, CA. UC Davis expects to
partner with the insurance company AAA Northern California, Nevada and Utah to secure
participants. UC Davis and AAA have previously negotiated agreements in which AAA screens
their policy holders and mails invitation to qualified people. AAA’s screening is fairly
rudimentary, involving minimum insurance coverage and zip code. More refined screening is
accomplished through a screening questionnaire hosted by UC Davis. From the point in time
when UC Davis contacts a volunteer who has completed this on-line screener, UC Davis is the
sole point of communication with the participant.
PHASE 2: FULL STUDY
The full study design will be refined based on results of the pilot study. The refined scope of
work for the full-study will include a list of cities in northern California and Nevada in which to
recruit volunteer participants, a final sample size goal, an estimate of the minimum population
difference in fuel economy that can be reliably estimated with such a sample size, a specific invehicle
device, a protocol for delivering, installing, operating, uninstalling and retrieving the in-vehicle
devices, etc. A rough outline of the current study design and objectives is presented
below.
Equipment Procurement
o UC Davis will purchase and configure an additional twenty-five (25) interface
units capable of displaying and recording fuel-economy related driving data and
GPS signals.
Sampling Protocol
o The full study design will include AAA based volunteer participant recruitment in
cities in northern California and Nevada. The sample will be refined by vehicle
type to insure that none of the study vehicles currently display fuel economy, and
that they are all compatible with the selected study interface. Volunteers will be
selected in each of the cities for an expected total sample of one hundred and fifty
volunteers.
Experimental Protocol
o The experiment will proceed in geographically and seasonally paired cohort
waves of approximately fifteen volunteers at one-month intervals. The cohort
pairs will be chosen to provide comparisons between seasons in the same region,
Fuel Economy Instrumentation Study
Attachment A
page 3 of 8
and between regions during periods of similar weather. Each 5 weeks, a new
cohort will begin the experiment, resulting in a total data collection period of 50
weeks.
o Each volunteer will participate for two months. During the first month, the
participant will drive with an active logging device, but a deactivated screen in
their vehicle. During the second month, the screen will be activated by the
researchers, and the driver will be able to view in-vehicle feedback about their
fuel economy.
o Each volunteer will complete an electronic survey (see survey protocol below)
two times in order to track changes in their attitudes and knowledge about fuel economy.
o Researchers will conduct semi-structured interviews of each participant both at
the beginning and at the end of the placement (see interview protocol below).
These interviews will provide insight into the participants understanding, interest,
and interpretations of the new fuel economy information.
Data analysis and Report
o UC Davis will analyze the data and prepare a report with the findings and experimental. In
particular, the report will address the research questions:
What is the distribution of real-world fuel economy increases that can be
expected over the long term from in-vehicle fuel economy displays?
Do the distributions differ by vehicle type?
City/region?
How many people know driving style matters to fuel economy?
What do they think the range of effects is?
What do they think other consequences are?
If the desired effect of in-vehicle fuel economy displays is something that
could be called “eco-driving,” then what proportion of the population are
already eco-driving?
What is the effect of IVDs on eco-drivers? On non-eco-drivers?
Fuel Economy Instrumentation Study
Attachment A
page 4 of 8
Survey Protocol
University of California, Davis Vehicle Study
This survey includes approximately thirty simple questions and should take less than
ten minutes to complete. This survey is intended for you to complete privately by
yourself, and each participating household member should complete the survey on
their own. Finally, please fill out the survey in the order shown without skipping any
questions.
We use the phrase “fuel economy” to mean the miles per gallon you get when driving
your car.
Please respond to the statements you read below by indicating your agreement or
disagreement with them.
Strongly
Disagree
1.
Driving fast is fun
2.
Saving gasoline makes me happy
3.
Driving efficiently is unsafe
4.
Driving is expensive
5.
Using gasoline lets me do what I need to do
6.
I like to know all about my car
7.
Saving gasoline is important
8.
I’m not about to change my driving habits
9.
I like to master new technologies
Neither
Disagree Agree nor
Disagree
10. Saving time is more important than saving gasoline
11. When driving it is best to "go with the flow"
12. Most people can save gasoline by driving more efficiently
13. I don't care if other drivers think I'm slow
14. It is important to drive at or below the posted speed limit
15. I have had bad experiences with other drivers getting mad at me
Fuel Economy Instrumentation Study
Attachment A
page 5 of 8
Agree
Strongly
Agree
Strongly
Disagree
Disagree
Neither
Agree nor
Disagree
Agree
Strongly
Agree
16. How I drive can change my car’s fuel economy
17. I know how to drive to get higher fuel economy
18. I already drive to get the highest fuel economy I can
19.
My fuel economy is a result of (too many) things beyond my personal
control
20. How I drive is determined by roads and traffic
For these next few questions, make your best guess if you aren’t sure.
Wild
Guess
21a.
I get
Very
Somewhat
Confident
Confident
mpg in my primary car
21b. How confident are you about that number?
22a. I could make an
changing my driving style.
% improvement in my my primary car fuel economy by
22b. How confident are you about that number?
For these next few questions, respond to the statements you read below by
indicating your agreement or disagreement with them.
Strongly
Disagree
Neither
Disagree Agree nor
Disagree
Agree
Strongly
Agree
23. I'm a perfectionist
24. I like to be the person in a group who has the right answer
25. I don't care if I win in competitions
26. I'll try something a second or third time to get it right
27. I usually leave things at "good enough"
For these final questions, please rank the top three items that are the most
important to you when driving, starting with the most important.
Not at
all
important
28a. When
driving,
I
would
most
like
Fuel Economy Instrumentation Study
Attachment A
page 6 of 8
Very
Somewhat
important
important
to
How important is this to you?
Secondly,
28b.
I
would
most
like
most
like
...
to
How important is this to you?
Thirdly,
28c.
I
would
...
to
How important is this to you?
Submit
Fuel Economy Instrumentation Study
Attachment A
page 7 of 8
Interview Protocol
All questions are open-ended. Total interview time is one-hour.
1.
Is increasing your fuel economy an important goal for you, compared to other things you accomplish by
driving?
a.
Why?
2.
What are the things a driver can do while driving their car to increase their fuel economy?
3.
Before we provided you with the fuel economy display, did you do any of these things? How many of these
things would you say you do all the time while you are driving? Most of the time? As often as not?
Seldom? Never?
a.
4.
After we provided you with the fuel economy display did you do any of these things? How many of these
things would you say you do all the time while you are driving? Most of the time? As often as not?
Seldom? Never?
a.
5.
For each of these, why do you do them as often or as seldom as you do?
For each of these, why do you do them as often or as seldom as you do?
Do you feel that your driving time is longer, or about the same when you do these things?
a.
How important is it to you to get around quickly?
b.
On a typical day, are you in a hurry? How does being in a hurry change your driving habits?
6.
Would you say that you found the display useful or useless? Informative? Distracting?
7.
If we could leave the display with you, would you use it, or would you rather we take it away?
8.
Imagine the next car you might buy. Would you want that car to display your fuel economy to you?
9.
a.
Would you want to turn that display off and on?
b.
Why?
c.
When would you turn it on? When would you turn it off?
As a passenger in the car, did you feel like you could notice changes in the drivers driving style? If yes, did
those changes make you more or less comfortable? Did those changes make you think the driver was
accomplishing an important goal?
10. Did you talk with friends, family, co-workers, or anyone else about your participation in this research? If
yes, what did you tell people about it? What was their response?
11. During the study period, did you substantially change the places you drive to or your routes?
a.
Why?
Fuel Economy Instrumentation Study
Attachment A
page 8 of 8
| File Type | application/pdf |
| File Modified | 2010-07-16 |
| File Created | 2010-07-16 |