Comparison of Owner Perceived and
EPA Measured Fuel Economy
James A. Rutherford
August 1977
Characterization and Applications Branch
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air & Waste Management
U.S. Environmental Protection Agency
-------
Abstract
Data from 1099 vehicles (model years 1974-1976) in the Fiscal Year 1975
Emission Factor Program are utilized in examining the differences
between owner estimated fuel economies and fuel economies derived from
EPA tests on the same in-use, consumer-owned vehicles. The discrepancies
are examined in terms of absolute differences and percentages. Various
vehicle classification, maintenance and utilization factors are investigated
to determine their relationship to these discrepancies. The agreement
in ranking of vehicles on fuel economy between owner determined and EPA
determined economies is also investigated.
-------
Background
The Gas Mileage Guide presented by EPA in conjunction with FEA is a tool
for comparing new cars on the basis of fuel economy. Test sequences
which produce the figures for the guide are precisely defined in an
effort to provide uniformity in evaluation and more scientifically
comparable results. Prototype vehicles at 4,000 miles are driven on a
dynamometer by professional drivers in a 75° F environment. Starting,
stopping, acceleration, and deceleration within the city and highway
cycles are intended to be representative of these modes of operation by
consumers.
It would not be expected that an owner calculating gas mileage for his
car would get the exact figure shown in the guide although the discrepancy
should not be too great. The difference between an owner determined gas
mileage and the guide value may be considered to contain two major
components. The first is the difference between the owner's determination
of gas mileage and the figures that would result if his car were put
through the test sequences used by EPA. These differences include the
specific type of driving, the ambient temperature, the vehicle engine
temperature, etc. The second is the difference between these tests run
on the consumer's in-use vehicle and the published figures in the guide
for that specific type of vehicle. These differences include prototype/
production differences as well as differences in specific vehicle
configuration such as axle ratio, test weight, tires, etc. This report
will focus upon the first of the two components.
Data
Data utilized in this report come from the Fiscal Year 1975 Emission
Factor Program. The program includes testing of 2200 vehicles from 1966
through 1976 model years. Consumer owned, in-use vehicles were selected
in seven cities based upon sales weighting for the determination of make
and model and based upon vehicle miles traveled for the determination of
model year characteristics. Information in this report is based upon
the model years 1974 through 1976.
City fuel economy is calculated for each vehicle from data obtained in
the 1975 Federal Test Procedure via the carbon balance method. Highway
fuel economy results were obtained on a subset of the vehicles via the
Federal Highway Fuel Economy Test. Various classification parameters
(e.g., engine size, transmission type, etc.) were recorded at reception
of the vehicle for testing. The vehicle owners were asked to complete a
questionaire which included information about vehicle use and maintenance
as well as the owner's estimates of city and highway fuel economies for
the vehicles. Of 1099 model year 1974 through 1976 vehicles included in
FY75 EFP, 239 had Highway Fuel Economy Tests performed. Owners gave
estimates of city fuel economy for 565 of the vehicles and estimates of
highway fuel economy for 539 of the vehicles. This resulted in 105
vehicles which had test values and owner estimates for highway fuel
economy and 565 which had both values for city fuel economy.
-------
Approach
In examining the discrepancies between test values and owner estimates
of fuel economy, the prima facie approach would be to consider one value
minus the other. However, a difference of 3 miles per gallon would
probably be more important when dealing with values around 8 miles per
gallon than around 30 miles per gallon. This would lead to a consideration
of some relative measure. In this report it is assumed that both absolute
and relative measures are of interest. The analyses are performed with
both types of measures thus providing the possibility of determining
whether they lead to consistent conclusions.
Absolute differences were calculated as owner's estimates minus test
values Ci-e-> owner's estimate of highway fuel economy minus the result
from HFET and owner's estimate of city fuel economy minus the result
from FTP). Relative measures were calculated as owner's estimate as
percent of test value (i.e., owner's estimate of highway fuel economy
divided by the result from HFET multiplied by 100 and owner's estimate
of city fuel economy divided by the result from FTP multiplied by 100).
Thus, if owner's estimate were less than the test value the difference
would be negative and the percent would be something less than 100, if
equal the difference would be zero and the percent 100, and if owner's
estimate were greater than the test value the difference would be
positive and the percent something greater than 100.
That the resultant differences were not all zero and that there was a
great deal of variability will be presented later. Beyond the overall
results it is of interest to determine whether the differences show any
systematic relationship to various vehicle classification and maintenance
factors. Due to the nature of the measurements being utilized, normal
theory statistical approaches do not seem appropriate. The non-parametric
method of choice for determining whether vehicle classification and
maintenance factors have significant statistical effects upon the
owner/EPA fuel economy differences and percents is the analysis of
variance test applied to ranks (termed the Kruskall-Wallis test).
Although it provides a test of significant factor effects, the Kruskall-
Wallis test is not amenable to determination of where the significant
differences occurr, i.e., the relationship among the levels of a factor.
For this purpose it is more enlightening to sort the vehicles into
meaningful groups and proceed with contingency table analysis. Since
the Kruskall-Wallis test does not provide easily interpretable information
and the contingency table analysis does not utilize as much of the
information contained in the data, both sets of analyses were performed
under the assumption that a comparison would be made to check for
consistency of conclusions.
In light of the stipulated purpose of the Fuel Economy Guide being the
comparison of vehicles, it is also of interest to consider the question
of how the ranking of the fuel economies of vehicles compares between
owners' estimates and test values. For this purpose a nonparametric
correlation measure is utilized. Due to the large number of tied
-------
observations the Goodman-Kruskal Gamma is used rather than the more
frequently seen Kendall's Tau. The G-K Gamma is similar to a normal
theory correlation coefficient in that its possible values range from
-1 to +1. As in normal theory a value of zero indicates independence
while values approaching +1 indicate strong agreement. From the G-K
Gamma an estimate of the probability of concordance is calculated where
the probability of concordance is defined as the probability that, for
two vehicles drawn at random from the appropriate stratum, if one of the
measures ranks one vehicle above the other then the other measure will
rank them in the same order.
Results
After calculation of owner's estimate as percent of test results,
vehicles were assigned to three equal (or nearly equal) sized groups
separately for city and highway fuel economies. A description of these
groups is presented in Tables 1-3. This independent grouping resulted
in quantiles with mean percentages of 79, 97, and 117 for city fuel
economy and quantiles with mean percentages of 74, 88, and 107 for
highway fuel economy. The range of city fuel economy percentages was
slightly larger (46-171) than the range of highway fuel economy percentages
(59-160). As seen in Table 3 the two groupings were not in close agreement.
For example, of 31 vehicles in the high group for highway only about
half (16) were vehicles in the high group for city fuel economy.
The differences (owner estimate minus test result) were used to group
the vehicles into the a priori categories: less than negative 2,
negative 2 to plus 2, and greater than 2. These city and highway
groupings are described in Tables 4-6. Sixty-nine percent of the owners
estimated within two miles per gallon of test results for city fuel
economy while 18 percent and 13 percent were more than two miles per
gallon respectively below and above test results. For highway fuel
economy 47 percent were within two miles per gallon of test results
while 45 percent were more than 2 belox^ and 8 percent were more than 2
above. The agreement between highway and city difference groupings
appears somewhat closer than percent groupings.
Tables 7 and 8 present comparisons of these groupings with classification
factors and questionnaire responses. A rough idea of the effect of the
factors can be obtained by looking down the mean columns in Table 7. If
the factor is unrelated to the percentages the mean percentages would be
expected to be equal across the levels of the factor. The figures
under the headings "Low", "Medium", and "High" in Table 7 and "< -2",
"-2 to +2" and "> +2" in Table 8 are counts of vehicles which fall
within that cell of the contingency table. These are the data upon
which the chi-square tests were performed. In Table 7, since the original
groups are of equal size, if the factor is unrelated to the percentages
the number of vehicles within the three groups should be about equal for
a row. Table 8 is harder to interpret. One would compare the distributions
among the rows of a factor. As will be seen later, the graphical
presentation of the cross-classifications which resulted in nominal
significance in Figures 1-19 is a much more rewarding approach to this
data.
-------
At this point it should be noted that absolute versus relative measures
and the chi-square test versus the Kruskall-Wallis for the most part
provided similar results. When owner vs test procedure fuel economies
were compared to identify any key classification variables, the discrepancies
among these approaches v/ere as follows: the presence of a catalyst was
nominally significant for both tests when performed on percents but for
neither when performed on differences; the questions relating to owner
satisfaction and frequency of tune-up were insignificant for the Kruskall-
Wallis test performed on differences while nominally significant on the
three other tests.
Tests Resulting in Nominal Significance
(* indicates significance at 0.01)
Percentages Differences
Chi-square Kruskal-Wallis Chi-square Kruskal-Wallis
Site * * * *
Model Year
Model Size * * * *
Cylinders * * * *
Carb Venturis * * * *
CID * * * *
Transmission * * * *
Manufacturer * * * *
Catalyst * *
Primary Use7
Maintenance7
Satisfaction? * * *
Often Tuned7 * * *
Last Tune?
Who Tuned?
Of the tests performed, one classification factor and several questions
from the questionnaire were found to be insignificant for all tests.
Model year was not significant and the questions relating to primary
use, maintenance according to manufacturer's specifications, time lapse
since last tune-up, and who performed the last tune-up were found to be
insignificant in relation to differences and percents. These questions
appeared on the questionnaire as follows:
1. How is this vehicle used7
a. Driver only,
b. Driver and 1 passenger,
c. Driver and 2 passengers,
d. Driver only with heavy cargo,
e. Driver, passenger and cargo,
f. Towing a trailer;
-------
Would you consider the vehicle has been maintained in accordance
with the manufacturer's recommendations7
a. Yes,
b. No,
c. Not sure,
d. Don't know;
How long ago was the last tune-up?
a. Too new not due,
b. Due but not done,
c. 0-6 mos.,
d. Over 1 yr.,
e. Don't know;
4. Who performed this tune-up?
a. No tune-up,
b. Dealer,
c. Independent garage,
d. Tune-up clinic,
e. Yourself,
f. Don't know.
The rest of the tests were nominally significant at the 0.01 level for
city fuel economies except as noted above. None of the tests were
significant for highway fuel economies. Whether the lack of significance
on highway fuel economies is due to the smaller sample sizes or due to a
true difference is not clear. Perusal of Tables 7 and 8 does not generally
indicate trend agreement across the factors for the city and highway
economies.
The significant tests are represented by Figures 1-19. Though complex
at first glance these figures can be very informative after some explanation
and provide a better overall picture of the Chi-square than can be
gleaned from the numbers presented in tabular form. For example, consider
Figure 3. The basic format of the figure is the same as a contingency
table. Vehicles are cross-classified by number of cylinders for rows
and by city percent group for columns. Within each resultant cell there
are two boxes and an angle. The solid box represents the actual observed
number of vehicles for that cell relative to the rest of the table. The
dashed-line box represents the expected value for the cell relative to
the rest of the table based upon the assumption of independence of rows
and columns. (The assumption of independence implies that knowing that
a vehicle belongs in a certain row of the table provides no information as
to which column of the table the vehicle is likely to belong. The expected
value for the cell is then the number of vehicles which would on the
average belong in the cell based upon this assumption.) Since this figure
is based upon percentage grouping for which the original groups were of
equal sizes the expected value for the A cylinder vehicles is the same
in each of the three percentage groups. The angle (measured counter-
clockwise from 3 o'clock) represents the cell's contribution to the chi-
square statistic. The sum of all the angles in the table is 360°.
Looking across the row for four cylinder vehicles it is seen that
-------
observed values are smaller than expected for low and medium city percent
groups while the observed value is larger than the expected for the high
group. This would indicate that four cylinder vehicles tend to have
owner estimated city fuel economy as a higher percentage of the test
value than would be expected under the assumption that the percentages
are independent of the number of cylinders. Analogously, six and eight
cylinder vehicles appear to have lower percentages than would be expected
under the independence assumption. The major contributions to the chi-
square come from the low and high percent groups of four cylinder vehicles.
In this manner the following relative inferences may be drawn with
respect to owner estimated city fuel economies relative to EPA test fuel
economies:
Site (Figures 1 and 11): At the two extremes vehicles from Phoenix
and Chicago have respectively high and low owner estimates
while the other five cities show less marked divergences from
the expected.
Model Size (Figures 2 and 12): Owners of subcompact (and to a
lesser extent compact) vehicles tend to estimate high relative
to test results.
Number of Cylinders (Figures 3 and 13): Four cylinder vehicles
tend to have high owner estimates while six and eight cylinder
vehicles have lower owner estimates.
Number of Carburetor Venturis (Figures 4 and 14): Vehicles with
one venturi tend to have low owner estimates while vehicles
with two have higher.
Engine Size (Figures 5 and 15): The vehicles with smallest engines
(0-150 CID) have high estimates while vehicles with moderate
engine sizes (331-399 CID) have low owner estimates.
Transmission Type (Figures 6 and 16): Automatics have low estimates
and manuals have high.
Manufacturer (Figures 7 and 17): The group of vehicles which are
not manufactured by AMC, Chrysler, Ford, or GM have high owner
estimates.
Presence of Catalyst (Figure 8): Owners of vehicles with catalysts
estimate low while owners of vehicles without catalysts tend
to estimate high.
-------
"Overall, Are You Reasonably Satisfied with the Engine Performance
of this Vehicle?" (Figures 9 and 18): Owners who gave the answer
"yes" to this question estimated high while owners who answered
"Most of the time" tended to estimate low.
"How Often is This Vehicle Tuned-up?" (Figures 10 and 19): Owners
who answered "No tune-up yet" tended to estimate low while
those who answered "Every 6 mos." tended to estimate high.
Tahle 9 presents the results of the calculations of the Goodman-Kruskal
Gamma and an estimate of the probability of concordance for appropriate
groups. These measure the agreement in ranking between owner estimates
and EPA tests. The Goodman-Kruskal Gamma is similar to a correlation
coefficient ranging from -1 to +1 and the probability of concordance is
the probability that two vehicles drawn at random from the appropriate
group would be ranked in the same order by the two determinations of
fuel economy. These groups were selected since they correspond to the
current organization of the Fuel Economy Guide and a breakdown which
would be practical for a consumer considering the purchase of a new
vehicle. It is seen that for larger groupings (all vehicles, all 1976
vehicles, etc.) the probabiltiy of concordance and the G-K Gamma is
larger than for the more discrete breakdowns. This is reflective of the
fact that the smaller groups are fairly homogeneous within while exhibiting
a large degree of heterogeneity amongst the groups.
Conclusions
Only about half of the owners who gave estimates of highway fuel economy
and had the Highway Fuel Economy Test performed on their vehicles were
within two miles per gallon of the test result while nearly half of them
estimated three or more miles per gallon less than the results of the
test. Similar calculations on city fuel economy show 69 percent within
two miles per gallon of test results with a reasonably comparable number
above and below (13% and 18% respectively).
Data used in this report show no statistically significant effect of
vehicle classification, maintenance, or use factors upon the discrepancies
between owner estimates and EPA test determinations of highway fuel
economy. However, (possibly due to the larger effective sample sizes)
many of these factors were found nominally significant for city fuel
economy discrepancies. Various trends were located within these factors
which indicate that significant differences might be attributable to
some psychological effects rather than the presumed technological
shortcomings of EPA test determinations of fuel economies. In terms of
classification factors, it is generally the vehicle which would be
expected to achieve high fuel economy for which the owner's estimate of
fuel economy is relatively high compared to the test result while vehicles
for which fuel economy would be expected to be mediocre or low show
relatively low estimates of fuel economy compared to test results. In a
similar vein, those owners who were relatively happy with engine performance
-------
and had the vehicle tuned-up at regular six month intervals had comparatively
high fuel economy estimates. On the other hand factors which would seem
more likely to influence the relationship between owners' estimates and
EPA test results such as model year and usual vehicle load did not show
up significant in the analysis.
-------
Table 1
City Fuel Economy Groups
Owner's Estimate as Percent of FTP Result
Highway
N
188
188
189
Minimum Maximum
46 89
90 104 ,
104 171
Table 2
Mean
79
97
117
Highway Fuel Economy Groups
Owner's Estimate as Percent of FET
N
35
35
35
Minimum Maximum
59 83
83 94
94 160
Table 3
Mean
74
88
107
Comparison of City and Highway
Fuel Economy Groupings
Low
Medium
High
City
Low Medium
21 7
13 6
4 11
High
1
11
16
-------
Table 4
City Fuel Economy Difference Groups
Description of FTP Results
Owner Estimate - FTP N % Min
Max
< -2 102 18 11 29
-2 to +2 387 69 8 30
> +2 71 13 9 28
Table 5
Highway Fuel Economy Difference Groups
Description of HFET Results
Owner Estimate - HFET N % Min Max
< -2 47 45 16 39
-2 to +2 50 48 13 35
> +2 8 7 16 30
Table 6
Comparison of City and Highway
Fuel Economy Groupings
Highway
< -2 -2 to +2
< -2 16
City -2 to +2 19
> +2 3
8
30
6
Mean
16
14
18
Mean
22
20
22
> +2
1
2
4
-------
Table 7
City and Highway Fuel Economy Percents
by Classifications and Questionnaire Response
City Highway
Site
Chicago
Denver
Houston
Los Angeles
St. Louis
Washington
Phoenix
Model Year
1974
1975
1976
Model Size
Full Size
Intermediate
Compact
Sub compact
Truck
Cylinders
4
6
8
N
164
39
47
69
52
60
134
155
153
257
142
126
108
142
44
134
78
351
Mean
92
96
103
100
101
95
102
99
97
98
99
95
97
103
87
103
92
97
Low
70
17
12
23
12
25
29
46
54
88
39
57
34
31
27
27
34
127
Medium
61
10
11
18
19
23
46
55
49
84
59
35
42
39
11
36
28
124
High
33
12
24
28
21
12
59
54
50
85
44
34
32
72
6
71
16
100
N
22
17
16
24
4
20
2
1
1
103
17
21
23
19
23
22
22
61
Mean
83'
88
88
96
93
90
102
104
102
89
91
87
92
91
86
92
95
87
Low
12
6
4
7
1
5
0
0
0
35
4
8
8
4
11
4
6
25
Medium
4
7
9
4
1
10
0
0
0
35
7
7
6
8
7
9
6
20
Higl
6
4
3
13
2
5
2
1
1
33
6
6
9
7
5
9
10
16
-------
Table 7 (con't)
City
Highway
Carb Venturis
1
2
4
Fuel Injection
CID
0-150
151-250
251-330
331-399
>_ 400
Transmission
Automatic
Manual
Manufacturer
AMC
Chrysler
Ford
GM
Other
N
87
328
133
14
136
60
90
163
116
439
126
21
73
121
248
102
Mean
93
99
97
111
104
93
95
95
99
96
104
96
95
98
95
106
Low
37
103
47
1
27
26
36
68
31
163
25
7
30
42
95
14
Medium
30
107
47
3
36
22
29
52
49
155
33
7
23
39
91
28
High
20
118
39
10
73
12
25
43
36
121
68
7
20
40
62
\
60
N
23
49
28
5
22
16
20
29
18
83
22
3
17
18
49
18
Mean
96
88
86
85
92
100
84
85
90
89
94
92
91
86
90
92
Low
6
17
11
1
4
3
11
13
4
32
3
1
6
6
19
3
Medium
6
18
10
1
9
4
5
9
8
26
9
1
5
8
13
8
Higl
11
14
7
3
9
9
4
7
6
25
10
1
6
4
17
7
-------
Table 7 (con't)
Highway
N Mean Low Medium High N Mean Low Medium High
90 89 33 27 30
15 91 2 85
Catalyst
Yes 344 96 130 117
No 221 100 58 71
Primary Use
Driver only 341 97 180 173
Driver & 15 99 3 8
1 Passenger
Driver & 19 100 5 6
2 Passengers
Maintained
According to
Mfg Rec.?
Yes 530 98 180 173
No 15 99 3 8
Not Sure 19 100 5 6
Satisfied
with Engine
Performance7
Yes 457 99 134 154
Most of 69 92 34 21
the t ime
97
92
177
4
8
177
4
8
166
14
66 90 24 19 23
21 89 7 77
13 90 3 64
104
0
1
90
72
34
0
1
35 35
0 0
0 0
83 90 28 28 27
15 91 5 4 6
No
39
90 17
13
87
-------
Table 7 (con't)
Cit5
Highway
How often
Tuned7
Not Yet
Mfg. Rec.
6 Months
Year
Less Often
Don ' t Know
Last Tune7
Too new
Due, not
done
0-6 months
6-12 months
Over 1 Year
Don't Know
Who Tuned7
None
Dealer
Ind. Garage
Clinic
Self
Don ' t Know
N
209
95
120
111
20
10
210
29
236
63
21
6
234
156
83
19
63
10
Mean
94
98
103
98
98
109
95
92
100
101
97
98
95
100
101
100
98
100
Low
87
34
27
32
8
0
83
11
70
14
8
2
93
44
25
8
15
3
Medium
70
27
39
44
4
4
69
11
77
23
6
2
77
47
28
4
27
5
High
52
34
54
35
8
8
58
7
89
26
7
2
64
65
30
7
21
2
N
76
14
8
7
0
0
78
6
20
1
0
0
84
15
2
0
4
0
Mean
88
96
91
95
—
—
88
87
95
102
—
—
88
96
89
—
96
Low
30
2
2
1
0
0
30
2
3
0
0
0
32
2
1
0
0
0
Medium
26
4
3
2
0
0
25
3
7
0
0
0
28
4
1
0
2
0
High
20
8
3
4
0
0
23
1
10
1
0
0
24
9
0
0
2
0
-------
Table 8
City and Highway Fuel Economy Differences
by Clasifications and Questionnaire Responses
Site
Chicago
Denver
Houston
Los Angeles
St. Louis
Washington
Phoenix
Model Year
1974
1975
1976
Model Size
Full Size
Intermediate
Compact
Subcompact
Truck
Cylinders
4
6
8
< -2
41
10
4
12
9
15
11
25
28
49
13
28
20
25
16
23
25
54
City
-2 to +2
114
24
33
46
36
37
97
109
106
172
118
87
76
77
27
71
47
268
> +2
8
5
10
10
6
8
24
20
18
33
11
9
11
38
1
3
6
26
< -2
13
7
7
8
2
10
0
0
0
47
4
9
10
11
13
11
9
27
Highway
-2 to +2
9
10
8
12
1
8
2
1
1
48
13
12
9
6
8
9
9
32
> +2
0
0
1
4
1
2
0
0
0
8
0
0
4
2
2
2
4
2
-------
Garb Venturis
1
2
4
Fuel Injection
CID
0-150
151-250
251-330
331-399
>_ 400
Transmission
Automatic
Manual
Manufacturer
AMC
Chrysler
Ford
GM
Other
< -2
29
53
20
0
23
17
22
30
10
82
20
4
17
21
47
13
Table
City
-2 to +2
49
222
105
8
72
37
58
122
98
318
69
16
47
86
184
54
8 (Con't)
> +2
8
50
8
5
39
6
9
9
8
37
34
1
9
11
16
34
< -2
9
23
13
2
11
6
11
15
4
38
9
1
8
7
22
9
Highway
-2 to +2
9
23
15
3
9
6
7
14
14
40
10
2
7
11
22
8
> +2
5
3
0
0
2
4
2
0
0
5
3
0
2
0
5
1
-------
Table 8 (Con't)
City
< -2 -2 to +2 > +2
Catalyst
Yes 69 237
No 33 150
Primary Use
Driver Only 68 228
Driver and 23 103
1 Passenger
Driver and 9 40
2 Passengers
Maintained
According to Mfg.
Recommendations?
Yes 98 362
No 1 12
'Not Sure 3 12
Satisfied with
Engine Performance?
Yes 72 314
Most of the 18 48
Time
No 12 25
+2
34
37
42
18
7
65
2
4
66
3
Highway
< -2 -2 to +2
41 41
6 9
29 31
10 10
6 6
46 50
0 0
1 0
36 41
8 5
> +2
8
0
6
1
1
8
0
0
6
2
-------
Table 8 (Con't)
City Highway
How Often
Tuned'
Not Yet
Mfg. Rec.
6 Months
Year
Less Often
Don ' t Know
Last Tune?
Too New
Due, Not Done
0-6 Months
6-12 Months
Over 1 Year
Don't Know
Who Tuned?
None
Dealer
Ind. Garage
Clinic
Self
Don't Know
< -2
48
23
10
17
4
0
42
9
38
8
4
1
51
26
13
2
9
1
-2 to +2
143
55
84
86
12
7
146
19
156
50
12
4
160
103
58
14
45
7
> +2
18
16
22
8
4
3
22
1
37
5
5
1
23
25
9
3
9
2
< -2
39
3
4
1
0
0
38
3
6
0
0
0
41
4
1
0
1
0
-2 to +2
32
8
4
6
0
0
35
3
11
1
0
0
38
9
1
0
2
0
> 4-2
5
3
0
0
0
0
5
0
3
0
0
0
5
2
0
0
1
0
-------
Table 9
The Goodman-Kruskal Gamma and the
Probability of Concordance
Highway
Cit}
Model
Year
all
1974
ii
ti
ii
ii
ii
1975
n
it
n
n
n
1976
n
n
ii
ii
n
Model
Size n
all 105
all
Full size
Intermediate
Compact
Sub compact
Truck
all
Full Size
Intermediate
Compact
Subcompact
Truck
all 103
Full Size 17
Intermediate 21
Compact 23
Subcompact 19
Truck 23
gamma j^ n
.5532 .7766 565
155
47
40
30
36
0
153
39
29
26
39
19
.5383 .7692 257
-.4433 .2784 56
.1123 .5562 57
.4123 .7062 52
.4783 .7392 67
.1739 .5870 25
gamma
.6418
.6237
.2895
.0541
.3577
.3425
.6446
.3028
.3850
.4698
.6190
.2740
.6310
.2670
.0830
.2784
.5581
.3655
£
.8209
.8119
.6448
.5271
.6789
.6713
.8223
.6514
.6925
.7349
.8095
.6370
.8155
.6335
.5415
.6392
.7791
.6828
-------
Figure 1
CITY FUEL ECONOMY PERCENT GROUP BY SITE
CHICAGO
IDS
DENVER
HOUSTON
bJ
ST.LOUIS
HRSHINGTON
PHOENIX
'JCJSG
LOW
MEDIUM
CITY PERCENT GROUP
CHI SaUflRED IS 42.038
HIGH
-------
Figure 2
CITY FUEL ECONOMY PERCENT GROUP BY MODEL
FULLSIZE
INTERMEDIRTE
u
tvl
i—i
en
COMPflCT
o
o
z:
SUBCOMPflCT
! |
L-
T
M«Mfll
I
TRUCK
ce-
LOW
MEDIUM
CITY PERCENT GROUP
HIGH
CHI SQUflRED IS 50.U33
-------
Figure 3
CITY FUEL ECONOMY PERCENT GROUP BY CYLINDER
FOUR
m
lp^
1
J
--C_- I
1 _!
(0
cc
UJ
o
^ SIX
o
EIGHT
L
1
-n
^
LOH
MEDIUM
CITY PERCENT GROUP
HIGH
CHI SQUflRED IS
-------
Figure 4
CITY FUEL ECONOMY PERCENT GROUP BY CURB
ONE
to
M
cc
a
t-
wTHO
> V-
cr
u
FOUR
LOH
NEOIUN
CITY PERCENT GROUP
HIGH
CHI SQUARED 19 6.B78
-------
Figure 5
CITY FUEL ECONOMY PERCENT GROUP BY CIO
0-150
i— • -T- —
I «2—1—I
L ' '_!
iL^
1 i
cc 151-250
a.
to
2 251-330
3C
O
« 331-339
01
u
r
~~I^r
i
!CH3i
C3
ESI
r
u
J—=3
1, ,„,., ii
LOW
CHI SQUARED IS 43.804
MEDIUM
CITY PERCENT GROUP
HIGH
-------
Figure 6
CITY FUEL ECONOMY PERCENT GROUP BY TRflNSHlSSlON
RUTOMflTIC
z
o
»••
to
to
to
z
-------
Figure 7
CITY FUEL ECONOMY PERCENT GROUP BY HflNUFflCTUREft
flHC
oc
tu
oc
o
CHRYSLER
FORD
6H
f — 1
1 — 1
1 1
L 1
1 — 1
i— — — —
• 1
L J
i
T~ ,,7
i
Lb^rz
~i i Ij^i.-**i
M^, |
1
: U
OTHER
-4Z
LOW
CHI SQUARED IS 42.763
MEDIUM
CITY PERCENT GROUP
HIGH
-------
Figure 8
CITY FUEL ECONOMY PERCENT GROUP BY CflTflLYST
YES
UJ
to
UJ
oc
v>
-------
Figure 9
tit
u
CITY FUEL ECONOMY PERCENT GROUP BY SflTISFflCTION
flC
o
u.
5 YES
o_
UJ
1
1
1
1
1
L
"1
1
1
1
e»
SS
u
r MOST OF TIME
o
UJ
•-••
u.
to
-------
Figure 10
CITY FUEL ECONOMY PERCENT GROUP BY TUNE
NOT YET
MFC REC
b_
O
3
O
6 MONTHS
I YEflR
l_
U- —
—J
LESS OFTEN
LOH
MEDIUM
CITY PERCENT GROUP
HIGH
CHI SQUflRED IS 22.580
-------
Figure 11
CITY FUEL ECONOMY DIFFERENCE GROUP BY SI TE
CHICflGO
DENVER
HOUSTON
U4
h-
tn L.fl.
ST.LOUIS
HflSHINGTON
PHOENIX
CHI SQUflRED IS 31.696
-2 TO +2
CITY DIFFERENCE GROUP
>*2
-------
Figure 12
CITY FUEL ECONOMY DIFFERENCE GROUP BY MODEL SIZE
FULL SIZE
r ~
i
L_
i
i
INTERHEDIflTE
IW
I— I
to
COMPflCT
o
o
z:
SUBCOMPflCT
TRUCK
CHI SQUflRED IS 57.331
-2 TO +2
CITY DIFFERENCE GROUP
-------
Figure 13
CITY FUEL ECONOMY DIFFERENCE GROUP BY CYLINDERS
to
tc
Ul
o
z
u
8
1
1
1
1
1
L
~1
_^J-
— •< i
i
1
I
-2 TO +2
CITY DIFFERENCE GROUP
CHI SQUflRED IS 54.469
-------
Figure 14
CITY FUEL ECONOMY DIFFERENCE GROUP BY CflRB VENTURIS
CflRB VENTURI
ro
G3
-2 TO +2
CITY DIFFERENCE GROUP
CHI SQURREO IS 23.519
-------
Figure 15
CITY FUEL ECONOMY DIFFERENCE GROUP BY CID
0-150
z
UJ
y:
U
a.
to
151-250
251-330
z:
o
z
331-399
CO
>399
CHI SQUflREO IS 60.425
-2 70 +2
CITY DIFFERENCE GROUP
-------
Figure 16
CITY FUEL ECONOMY DIFFERENCE GROUP BY TR flNSMISSION
u» RUTOMflTIC
a.
to
to
i-«
z:
to
•z.
cc
cc
MflNUflL
-2 TO +2
CITY DIFFERENCE GROUP
CHI SaUflREO IS 33,027
-------
Figure 17
CITY FUEL ECONOMY DIFFERENCE GROUP BY Mft NUFflCTURER
RMC
CC
UJ
CC
=>
h-
O
CHRYSLER
FORD
=3
Z
CC
GM
OTHER
CHI SQURREO IS 52.633
-2 TO +2
CITY DIFFERENCE GROUP
-------
Figure 18
CITY FUEL ECONOMY PERCENT GROUP BY SflTISFflCTION
UJ
u
z
1 YES
o
li-
CC
UJ
0-
UJ
z
MOST OF THE TJHE
o
UJ
(O
- NO
cc
V)
-2 TO +2
CITY DIFFERENCE GROUP
CHI SQURRED IS 14.020
-------
Figure 19
CITY FUEL ECONOMY DIFFERENCE GROUP BY TU NE
NOT YET
Lrfr
o
1U
z
=> MFG. REC.
z
UJ
6 MONTHS
YEflR
IfiBI
CHI SQUflRED IS 24.315
-2 TO +2
CITY DIFFERENCE GROUP
------- |