EVALUATION OF THE CRASH EFFECTS OF
THE CHANGES IN SPEED ZONES IN VICTORIA
IMPLEMENTED DURING 1992-1993
(EXCLUDING 100 to 110km/h):
UPDATE INCLUDING 1990-1997 CRASH DATA
by
Stuart Newstead
& Sanjeev Narayan
July 1998
Report No. 136
11
MONASH UNIVERSITY ACCIDENT RESEARCH CENTRE
MONASH UNIVERSITY ACCIDENT RESEARCH CENTRE
REPORT DOCUMENTATION PAGE
Report No.
Report Date
136
July 1998
ISBN
07326 1434 1
Pages
26 + Appendices
Title and sub-title:
Evaluation of the crash effects of the changes in speed zones in Victoria implemented during 199293 (excluding 100 to 11 Okm/h) : update including 1990-1997 crash data.
Author(s)
Newstead, S. & Narayan, S.
Type of Report & Period Covered
General, 1989-1995
Sponsoring Organisation - This project was funded through a Research Contract from VicRoads,
Road Safety Department. 60 Denmark Street, Kew, VIC. 3101
Abstract:
During late 1992 and early 1993, a rationalisation of speed limits on Victorian roads was
undertaken in order to achieve credible speed limits which were uniform with the rest of Australia.
As part of this rationalisation, many speed zoning changes occurred across Victoria, with some of
the most notable being the phasing out of 75 km/h speed zones and the introduction of 50, 70 and
80 km/h zones. Under the rationalisation, posted speed limits on some road sections were increased
while on other road sections the posted speed limits were decreased.
This study updates an earlier one which evaluated the casualty crash effects of the speed zone
changes implemented in Victoria for speed zone changes other than 100km/h to 110km/h. Effects
are estimated for the program of speed zone changes as a whole as well as for each particular type
of speed zone change. Results are presented for the whole of Victoria as well as for metropolitan
Melbourne and the rest of Victoria separately. Estimated effects of speed zone changes on casualty
crash frequency are further related to changes in crash type as well as results of speed monitoring.
Results of analysis in metropolitan Melbourne showed an overall increase in casualty crash
frequency of 4.7% with marginal statistical significance whilst no statistically significant change in
casualty crash frequency was found in the rest of Victoria. The net effect of the speed zone changes
over Victoria as a whole was a statistically significant increase in overall casualty crash frequency
of 5.4%.
Recommendations for further research are made.
Key Words: (IRRD except where marked*)
speed limit, evaluation, injury accident, statistical analysis, traffic regulations, accident type,
research report
Reproduction of this page is authorised.
Ev ALUA nON
OF THE CRASH EFFECTS
OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993
III
ACKNOWLEDGMENTS
The authors wish to thank the VicRoads project staff, Sam Pirrotta, Patricia Liew and Pat Rogerson
for their helpful comments in overseeing the progress of this project. Thanks are also extended to
the staff of the VicRoads' Land Information and Surveys Department, GIS Section, and in
particular Helen Lau for her expert advice on using ARCVIEW. Michael Fitzharris of the Monash
University Accident Research Centre is acknowledged for his assistance in carrying out the GIS
component of the project including map preparation.
The authors also wish to acknowledge the expert guidance and useful comments of Mr Max
Cameron and Mr Bruce Corben, Senior Research Fellows at Monash University Accident Research
Centre, in the course of carrying out the project.
IV
MONASH UNIVERSITY ACCIDENT RESEARCH CENTRE
TABLE OF CONTENTS
1.0
INTR
1.1
1.2
2.0
0DUeTI 0N
1
BACKGROUND
AIMS OF THE EVALUA TION
1
2
STuny DESIGN
2.1
2.2
2.3
2.4
3.0
2
EVALUATION METHOD
SITE SAMPLING
ANALYSIS STRATIFICATION
HYPOTHESES TESTED
2
3
4
5
DATA ..............................•..•................•••...................................................................•...•..•.......
5
3.1
3.2
SPEED ZONE CHANGES
CRASH DATA
3.2.1
3.2.2
4.0
5
7
Crash database
Study periods
7
8
ANALYSIS METHODS
4.1
4.2
SITE IDENTIFICATION
CRASH DATA EXTRACTION
4.2.1
4.2.2
4.3
4.4
5.0
11
12
TS
12
METROPOLITAN MELBOURNE
12
Crash Frequency
Post Hoc Analyses
12
15
REST OF VICTORIA
5.2.1
5.2.2
5.3
9
9
STATISTICAL ANALYSIS METHODS
POST HOC ANALYSIS
5.1.1
5.1.2
5.2
9
9
Treatment Crash Data
Control Crash Data
RE S UL
5.1
9
20
Crash Frequency
Post Hoc Analyses
20
21
21
WHOLE OF VICTORIA
6. 0
DISCUS SI ON
22
7.0
CON CLUSI 0NS
24
8.0
RECO MMEND ATI NS
25
9.0
RE FE REN CES
26
0
EVALUATION
OF THE CRASH EFFECTS
OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993
v
LIST OF TABLES
TABLE 1 : Summary of speed zone changes (from report of the Parliamentary Inquiry into the
Revision of Speed Limits)
1
TABLE 2 : Percentage of treated sites to be sampled to allow statistically significant detection of
given treatment effects
3
TABLE 3 : Summary of the relevant information on speed zone changes supplied by each
VicRoads region
,
6
TABLE 4 : Metropolitan LGAs selected for evaluation of speed zone changes
6
TABLE 5 : Rural LGAs selected for evaluation of speed zone changes
7
TABLE 6 : Speed zone changes and proposed control matching speed zones of crashes in the
same LGA
10
TABLE 7 : Results of crash frequency analysis - Metropolitan Melbourne
13
TABLE 8 : Summary of speed monitoring results presented in the Parliamentary Inquiry report
and related casualty crash frequency analysis results. Metropolitan Melbourne
19
TABLE 9: Results of crash frequency analysis - rest of Victoria
20
TABLE 10 : Results of crash frequency analysis - All of Victoria
22
LIST OF FIGURES
FIGURE 1 : 2x2 crash data contingency table for a single treated site and its control
11
FIGURE 2 : 2x2 crash data contingency table for L treated sites and controls
11
FIGURE 3 : Percentage of DCA groupings for crashes occurring in the treatment areas of
metropolitan Melbourne before and after speed zone changes: All speed zone changes
15
FIGURE 4: Percentage of DCA groupings for crashes occurring in the treatment areas of
metropolitan Melbourne before and after speed zone changes: 100-80km/h speed zone change. 16
FIGURE 5 : Percentage of DCA groupings for crashes occurring in the treatment areas of
metropolitan Melbourne before and after speed zone changes :75-80km/h speed zone change .... 17
FIGURE 6: Percentage of DCA groupings for crashes occurring in the treatment areas of
metropolitan Melbourne before and after speed zone changes :75-60km/h speed zone change .... 17
FIGURE 7 : Percentage of DCA groupings for crashes occurring in the treatment areas of the
rest of Victoria before and after speed zone changes: All speed zone changes
21
LIST OF APPENDICES
APPENDIX A : Sample of speed zone change sites used in analysis: Metropolitan Melbourne
APPENDIX B : Sample of speed zone change sites used in analysis: Rest of Victoria
APPENDIX C : Listing of speed zone change sites evaluated in metropolitan Melbourne, length
of road covered by each speed zone change and estimated crash effects by speed zone change
type within LGAs
APPENDIX D: Listing of speed zone change sites evaluated in the rest of Victoria, length of
road covered by each speed zone change and estimated crash effects by speed zone change type
withinLGAs
APPENDIX E : Chart of definitions for classifying accidents (DCAs)
VI
MONASH UNIVERSITY ACCIDENT RESEARCH CENTRE
EXECUTIVE SUMMARY
During late 1992 and early 1993, a rationalisation of speed limits on Victorian roads was
undertaken in order to achieve credible speed limits which were uniform with the rest of Australia.
As part of this rationalisation, many speed zoning changes occurred across Victoria, with some of
the most notable being the phasing out of 75 kmIh speed zones and the introduction of 50, 70 and
80 kmIh zones. Under the rationalisation, new criteria for speed zoning of roads was developed
resulting in posted speed limits on some road sections being increased whilst on other road sections
the posted speed limits were decreased. This study evaluates the casualty crash effects of the speed
zone changes implemented in Victoria for all speed zone changes other than 100 kmIh to 110 km/h.
A pseudo experimental study design was used for the evaluation, examining changes in casualty
crash frequency before and after speed zone changes. Only a sample of sites that had undergone
speed zone changes was used in the analysis. The analysis also incorporated the use of control sites
to represent parallel changes in casualty crash frequency due to other factors. Control crashes were
selected from all roads where speed zone was unchanged and matched with treatment crashes by
local government area and level of urbanisation. Crash data from three years before and three years
after implementation of the speed zone changes was analysed.
Analysis of the effects of speed zone changes on casualty crash frequency in metropolitan
Melbourne showed an overall increase in casualty crash frequency of 4.7%, although this result was
of marginal statistical significance and should be interpreted with caution. This estimated increase
represents in the order of 235 extra casualty crashes per annum across Melbourne due to all speed
zone changes. Assessment of the general effects on casualty crash frequency of increasing zoned
speed or decreasing zoned speed showed no statistically significant change in casualty crash
frequency when the zone speed was decreased, and a 9.3% casualty crash increase (with high
statistical significance) when zone speed was increased. For particular speed zone changes, the
change from 100 to 80 kmIh showed a highly statistically significant casualty crash reduction of
46%, translating to a saving of approximately 44 casualty crashes per annum across Melbourne.
Also the change from 75 to 60 kmIh showed a marginal statistical significance casualty crash
increase of 43%, representing an increase of around 151 casualty crashes per annum. Increased
speed zoning from 75 to 80 kmIh showed a highly statistically significant casualty crash frequency
increase of 10.7%, representing an increase of approximately 188 casualty crashes per annum
across Melbourne.
The results of analysis of casualty crash frequency in metropolitan Melbourne were generally
consistent with the results of speed monitoring.
Most of the speed zone changes which occurred in the rest of Victoria took place on the fringes of
country towns in the speed transition zones between 100 kmIh zones of the open highway and 60
km/h zones of the built up town area. Analysis of the overall casualty crash frequency change for
all speed zone changes combined in the rest of Victoria as well as analysis by specific speed zone
changes showed no statistically significant changes in crash frequency.
The net effect of the speed zone evaluation over Victoria as a whole was a statistically significant
increase in overall crash frequency of 5.4%, although this result is largely driven by the results in
metropolitan Melbourne.
Ev ALUA nON
OF THE CRASH EFFECTS OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993
VII
viii
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
EVALUATION OF THE CRASH EFFECTS OF
THE CHANGES IN SPEED ZONES IN VICTORIA IMPLEMENTED DURING 1992-1993
(EXCLUDING 100 to 11 Okm/h):
UPDATE INCLUDING 1990-1997 CRASH DATA
1.0
INTRODUCTION
1.1 BACKGROUND
During late 1992 and early 1993, a rationalisation of speed limits on Victorian roads was
undertaken in order to achieve credible speed limits which were uniform with the rest of Australia.
As part of this rationalisation, many speed zoning changes occurred across Victoria, with some of
the most notable being the phasing out of 75 km/h speed zones and the introduction of 50, 70 and
80 km/h zones. Under the rationalisation, new criteria for speed zoning of roads was developed
resulting in posted speed limits on some road sections being increased whilst on other road sections
the posted speed limits were decreased.
Table 1 from the Parliamentary Inquiry into the Revision of Speed Limits (Road Safety Committee,
1995) details the speed zone changes which were implemented under the rationalisation and the
lengths of road to which they apply.
TABLE 1 : Summary of speed zone changes (from report of the Parliamentary Inquiry into the
Revision of Speed Limits).
Previous Limit 2360
70
80
Revised
Limit
2910
415
70
110
50
80
1460
435
50
550 (km)
- Increase
Length
Length
(kmlh)
-Decrease
ofofRoads
Roads
(km)
150
5
350
25
10
nd Total
It should be noted that Table 1 includes the speed zone change from 100 to 11Okm/h affecting some
435km of road. This particular speed zone change was not part of this study brief and will not be
considered further in this study.
According to Table 1 and the Parliamentary Inquiry (Road Safety Committee, 1995), the total
length of road affected by these changes considered in this study was 2475 km, or 9.3% of the total
arterial road system. Of this road length, 1925 km saw a speed zoning increase whilst 550 km saw a
speed zoning decrease. Of the total length with changed speed zone, 1810 km (94%), changed by 5
km/h, whilst the remaining 6% changed by between 10 and 30 km/h.
Further information contained in the Parliamentary Road Safety Committee's report indicates that
50% of the urban arterial road network has undergone a speed change, compared with only 3% of
the rural arterial network. Allowing for the relative lengths of the urban and rural networks, this
represents approximately 482 km of rural road and 1993 km of metropolitan road experiencing a
speed zoning change.
Results of an evaluation of the crash effects of the speed zone changes in Victoria are reported in
Newstead and Mullan 1996. This study was carried out relatively soon after the speed zone changes
were affected and includes in the evaluation only 12 to 18 months crash data post speed zone
changes. Consequently, many of the results from this initial study were inconclusive prompting the
authors to recommend that the evaluation be carried out again at a later date examining crash trends
over a longer time period after implementation of the speed zone changes
1.2
AIMS OF THE EVALUATION
This evaluation study aims to update the earlier study of Newstead and Mullan to include crash
data from a longer time period after implementation of the speed zone changes in Victoria. The
basic objective of the study is to determine whether the speed zone changes implemented in
Victoria, other than 100 km/h to 110 km/h, have affected casualty crash frequencies on the road
segments to which they apply. Analysis has centred on comparing the accident frequency before
and after implementation of the speed zone changes.
A secondary aim of the study is to determine, on those road segments where a significant casualty
crash frequency change was observed due to speed zone changes, the particular casualty crash types
which have been responsible for this change. This may enable the possible mechanisms leading to
the observed changes to be inferred.
2.0
STUDY DESIGN
2.1
EVALUATION METHOD
The primary aim of the evaluation is to assess the change in crash frequency attributable to the
rationalisation of speed zones. It is not, however, sufficient just to compare crash frequency in the
periods before and after speed zone changes to determine the effects of the change. This is because,
during the period of implementation of the speed zone changes, a number of other major road
safety campaigns have been under way in Victoria. Along with changes in social and economic
conditions, these have had a large impact on road trauma in the state (see Newstead et aI, 1995, for
a description of some of these programs and an estimate of their effects on casualty crash
frequency). Any attempt to measure the effects of speed zone changes on crash frequency must also
take into account changes due to other programs or influences.
Use of the casualty crash history at a set of control sites in the analysis allows for adjustment of
other time varying factors which may have affected crash frequency, such as changes in economic
factors or operation of other road safety programs. Appropriately chosen control groups will
provide a measure of the crash frequency changes associated with these other factors, leaving any
further changes associated with the speed zone rationalisation alone. The measure of the effect of
speed zone changes on crash frequency is made by comparing the before and after crash
frequencies at the treated sites adjusting for the parallel changes in crash frequency at the control
2
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
sites from the corresponding before and after treatment time periods. This evaluation format is
known as a quasi-experimental design as it follows the format of a fully randomised treatmentcontrol type experiment but differs in that the treatment sites are not chosen at random. In the
context of this study, a "treated site" constitutes a road length which has undergone a speed zone
change.
The quasi-experimental study design using treatment and controls is often used in the evaluation of
the effectiveness of accident black-spot treatments (Corben et al 1990, BTCE 1993). One issue
which often arises as part of these studies is that of regression- to-the-mean which can be a problem
in analysis when treatment sites are selected on the basis of high accident frequency as in the case
of accident black-spot treatments. Regression to the mean should not be an issue in this study as
sites chosen for speed zone changes were not generally chosen on the basis of accident history.
2.2
SITE SAMPLING
The program of speed zone changes in Victoria has involved rezoning a large number of discrete
lengths of road. Given the size and scope of the project, it was considered impractical in the initial
evaluation by Newstead and Mullan (1996) to include every length of road on which a speed zone
change had occurred because ofthe large number of these. It was therefore decided that a sample of
the sites with speed zone changes be taken for analysis, with the view that the results obtained from
analysis of the sample are representative of all speed zone changes in Victoria. This study uses the
same sampling frame as Newstead and Mullan (1996). Whilst the statistical power calculations
below are those derived by Newstead and Mullan (1996) based on one years' after treatment data,
they give the rationale for the sampling frame used. The extra data available in this study served to
increase the effective sampling frame in Melbourne and the rest of Victoria by two and three times
respectively, giving corresponding increases in statistical power.
Choice of an appropriate sampling frequency was critical for the efficiency and accuracy of the
original study as sample size is related to the statistical power of the analysis. Statistical power of
an analysis determines the minimum sample size required to detect a statistically significant effect
of a given magnitude. Power calculations for sampling frequency have been calculated for this
project and are detailed in Table 2 based on the assumption that one year's after treatment data will
be available. The power calculations shown assume the analysis methods detailed below will be
employed.
TABLE 2 : Percentage of treated sites to be sampled to allow statistically significant detection of
given treatment effects.
Area
10%
>100%
25%
55%
15%
20%
7%
Minimum
overall
10%
length)
sites
Percentage
to10%
becrash
sampled
of treated
(by
15%
40%
20% detected
30%
20%
requency
change to be
EVALUATION
OF THE CRASH EFFECTS
OF THE CHANGES IN SPEED ZONES IN VICTORIA
DURING
1992-1993 3
From preliminary inspection of the quantity of treatment site data it was considered appropriate to
sample 10% of the treated sites in the metropolitan area given the original project size. Table 2
shows that this would be able to reliably detect a minimum 15% percent change in overall crash
frequency at the treated sites in total.
Table 2 highlights potential problems in the analysis of speed zone changes in the rest of Victoria.
It is evident from Table 2 that, even sampling 100% of the treated sites in this area, the smallest
crash frequency change which could be reliably detected is somewhat greater than 10%. This is
caused by the relatively small percentage of roads in this region with changed speed zone combined
with the low accident rate per kilometre on these roads. Even to reliably detect a minimum 20%
crash frequency change, the original study would have required 55% of the treated sites to be
sampled. This sampling frequency was inconsistent with the magnitude of the original project
given the number and nature of treated sites in the rest of Victoria (many of the treated sites in the
rest of Victoria are in buffer zones between open highways and built up rural towns, which can
typically be as short as 600m). Considering this, it was decided to sample 15% of the treated sites
in the rest of Victoria, enabling a minimum 40% change in crash frequency to be reliably detected
in the original study.
Given the extra post treatment data available in this study, the sampling frames chosen would allow
minimum crash reductions of around 10% and 25% to be detected in Melbourne and the rest of
Victoria respectively.
2.3
ANALYSIS STRATIFICATION
To provide maximum detail on the effects of speed zone changes on crash frequency, the analysis
has been graduated into a number of levels. The analysis stratification levels to be used, from
coarsest to finest, are;
I. All speed zone changes in all LGAs aggregated : This is an overall measure of the
influence of the speed zone rationalisation program on casualty crash frequency.
2. Speed zones where speed limits increased and speed zones where speed limits decreased:
All the speed zone changes are categorised into one of two broad categories; speed limit
increase and speed limit decrease. This level of analysis measures the broad average
casualty crash effects of the speed limit increases and decreases that were undertaken.
3. Each individual type of speed zone change: This level of analysis measures the average
casualty crash frequency change due to each type of speed zone change (eg. 75-70 km/h,
60-70 km/h) across all road lengths with each particular speed zone change.
4. Each individual type of speed zone change within each individual LGAs : This is similar
to 3 above but measures any differences in speed zone change effects for each type of
speed zone change between one LGA and another.
It was not certain whether the quantity of data available for analysis would allow the assessment of
graduation level 4. It is, in theory, possible to proceed to one finer level of analysis, being the
assessment of crash frequency change on each individual road length which experienced a speed
zone change. Given the amount of crash data available, this finest level of analysis assessing speed
4
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
zone changes at individual sites was not considered likely to produce conclusive results and hence
not attempted.
Because of the known differences between roads and crash patterns in metropolitan Melbourne and
the rest of Victoria the analysis has been performed at each of the 4 graduated levels for
metropolitan Melbourne and the rest of Victoria separately as well as for Victoria as a whole.
2.4
HYPOTHESES TESTED
The aim of the statistical analysis undertaken in this study was to determine whether speed zone
changes had significantly influenced casualty crash frequency. Hence, the global null hypothesis
being tested in all the analyses presented is that of no casualty crash frequency change due to the
program of speed zone rationalisation. In formulating an alternative hypothesis for the statistical
test procedure, there is no clear a-priori reason to expect speed zone changes may increase or
decrease casualty crash frequency for either speed increases or decreases. This is because crash risk
is not merely a function of absolute traffic speed but also factors such as variance in vehicle speeds
within a traffic flow. Consequently, a two tailed alternative hypothesis was appropriate to use for
determining the statistical significance levels of the tests presented here. This statistical test
structure has been used in hypothesis testing at all the levels of analysis discussed in section 2.3
above.
3.0
DATA
3.1
SPEED ZONE CHANGES
Each VicRoads regional office in Victoria prepared a summary of speed zone changes in their
region that had been undertaken as part of the rationalisation program. This information was
supplied to MUARC for the evaluation in hard copy format. In order to accurately determine the
type, location and exact timing of each speed zone change in Victoria, the following information
was required in the data supplied:
1. Municipality of change (LGA)
2. Location start and end points on the road of each speed zone change
3. Speed zoning before and after the change
4. Date of speed zone change
Examination of the data supplied revealed no uniform or consistent method of recording such
details. Of the seven VicRoads regions, only the Northern and Eastern Regions were able to supply
all relevant details within the one spreadsheet in a readily useable form. For three of the remaining
five VicRoads regions, the required information could, in most cases, be obtained by manual
processing of the supplied information. Manual processing of these cases generally consisted of
determining zone change start and end points from details of speed sign placement on the road. For
the remaining two VicRoads regions (Western and North-East Regions), information supplied was
incomplete and could not be used for the purpose of this study.
Contact with the VicRoads regional offices was made in an attempt to obtain the required missing
data items. Apart from limited assistance with some implementation dates in certain regions, in
Ev ALUA nON
OF THE CRASH EFFECTS OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993 5
general the VicRoads regional offices were unable to supply in a timely manner the missing
information required to make the data useable in the evaluation. Table 3 summarises the useable
information supplied by each VicRoads region.
TABLE 3 : Summary of the relevant information on speed zone changes supplied by each
VicRoads region.
Location
Local
.t.t.t.tDate
VicRoads Region
.t.tsome
some
Implementation
(Start
and
Changes
End.tGovt. .t
Speed
Zone
For the LGAs within the metropolitan VicRoads regions where the installation date of the speed
zone change was available, date of installation was often only given within a six month period. In
order to maximise the amount of after-treatment crash data available, it was decided to restrict
analysis to those Local Government Areas that had speed zone changes between July 1993, the
beginning of the program implementation, and December 1993. The majority of metropolitan
LGAs had speed zone change implementation dates falling within this time frame. Most of those
with implementation dates falling in the time frame and with all the required information supplied
by VicRoads were used in the evaluation. Table 4 lists those LGAs satisfying the inclusion criteria.
For the purpose of this analysis, the LGAs used were those which existed in Victoria prior to the
progressive restructuring ofLGAs which began in 1993.
TABLE 4 : Metropolitan LGAs selected for evaluation of speed zone changes
Berwick East Metropolitan
Richmond
Bulla
Eltham
Chelsea
Melbourne
VicRoads
Keilor
SunshineWest
North
Diamond
Valley
Collingwood
Metropolitan
South
Preston
and Templestowe
Table 4 shows the spread of metropolitan Melbourne LGAs included in the analysis, ranging from
inner city LGAs, such as Camberwell, Collingwood and Melbourne, to urban fringe LGAs such as
Diamond Valley, Bulla and Cranbourne. A map of the metropolitan LGAs used in the analysis can
be found in Appendix A.
6
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
The three rural VicRoads regions which had supplied sufficient site data for use in the evaluation
were further examined to reveal that the implementation dates of the speed zone changes occurred
from August 1993 through to December 1994, with most occurring in the time frame January 1994
to December 1994. In order to have the maximum after-treatment crash data for analysis, it was
decided to only consider those regions that had speed zone changes from January to June 1994.
Analysis then centred on all rural LGAs with sufficient site description provided by VicRoads that
had speed zone changes implemented from January to June 1994. Table 5 lists those rural LGAs
meeting these criteria.
Table 5 shows the coverage of the selected LGAs across each of the three rural VicRoads regions
included in the analysis. The selected LGAs include a mix of large population centres, such as Sale
and Geelong West, as well as largely rural LGAs, such as Maffra and Otway. A map of Victoria
showing the LGAs included in the analysis for the rest of Victoria is given in Appendix B.
TABLE 5 : Rural LGAs selected for evaluation of speed zone changes.
MtRouse
Bet
Gisbome
Bet Eastern
Maffra
Mortlake
Wannon
South
Western
VicRoads Region
Strathfieldsaye
Heywood
Hampden
Glenelg
Camperdown
LeighNorthern
Port
Fairy
Alberton
Sale
Bass Island
Phillip
Wonthaggi
Barrabool
Queenscliff
Heytesbury
Otway
Geelong West
Analysis of the length of road with speed zone changes in those LGAs selected in each of
Melbourne and the rest of Victoria showed sufficient coverage to meet the statistical analysis power
requirements discussed in section 2.2 above.
A list of all the speed zone change sites included in the evaluation by LGA for metropolitan
Melbourne is given in Appendix C, along with the approximate length of road covered under each
speed zone change in each LGA and in metropolitan Melbourne as a whole. Analogous information
is given for the rest of Victoria an Appendix D.
3.2
CRASH DATA
3.2.1
Crash database
The crash database used in the analysis was the VicRoads database of Police reported casualty
accidents in Victoria. Analysis centred on data from the complete years 1990 to 1996. Due to the
timing of the study, the 1997 crash data for the complete year was not available. Discussions with
VicRoads revealed crash data for 1997 to be available from January to about October, with the data
being complete to June 1997. Hence a preliminary 1997 crash data file was obtained for the
Ev ALUA nON
OF THE CRASH EFFECTS OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993 7
analysis with data from this file being used for the period January to June 1997 meaning the total
period of data available was January 1990 to June 1997.
The VicRoads Police reported crash database records many variables describing crash details
(Green 1991). Of those available, the following were relevant to this study;
• Road Reference
intersections
Point (RRP) : describes crash location for crashes occurring
at road
• Road Segment (RS) : describes crash location for crashes occurring away from intersections
• Date of crash
• Recorded speed zone of road on which crash occurred (as recorded by the Victoria Police)
• Definition for Classifying Accidents (DCA) describing crash type
• Crash severity (fatal, serious injury or other injury)
For those LGAs listed in section 3.1 there was a total of 44,279 crashes over the period 1990 to
June 1997, almost double the number used in the previous study by Newstead and Mullan (1996).
3.2.2 Study periods
As described in section 2.1 above, the study design which was chosen for this evaluation is a
pseudo-experimental design utilising treatment and control sites. Under this study design, suitable
periods before and after implementation of the speed zone changes must be defined on which to
base the analysis. These periods must be defined with reference to the available crash data (January
1990 to June 1997) and the treatment implementation dates.
As described in section 3.1, speed zone changes in the sampled metropolitan Melbourne LGAs
were implemented over the period July to December 1993. It was decided to define the study
"before" period as July 1990 to June 1993 and the study "after" period as July 1994 to June 1997.
In rural Victoria, speed zone changes for those LGAs sampled occurred over the period July 1993
to June 1994. For these LGAs the before period was July 1990 to June 1993 and the after period
was July 1994 to June 1997. This provided crash data spanning a period of 36 months before and
after treatment for both the metropolitan Melbourne and the rest of Victoria analysis.
These choices of before and after period offer a number of benefits to both the metropolitan
Melbourne and rest of Victoria analyses. Firstly the before and after periods were equal length
hence providing a balanced analysis. Secondly, the before and after periods cover the same
calendar months eliminating possible confounding effects of seasonal bias which are known to
exist in the crash data. Finally, the periods chosen make maximum use of the available after
treatment crash data hence maximising the power of the analysis for the available data. The choice
of identical before and after treatment periods for every treated site in the analysis is also
convenient for data extraction and manipulation.
8
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
4.0
ANALYSIS METHODS
4.1
SITE IDENTIFICATION
As VicRoads were not able to supply a specific file of crash data at treated sites for use in the
project, one critical task involved in the project was the identification of crashes at treated sites.
After selection of the sample of treated sites to include in the analysis, the appropriate accidents
were identified using a Geographical Information System (GIS) computer package in conjunction
with crash location data supplied by VicRoads.
The desktop mapping package Arc View 3.0 was used to identify sites with speed zone changes in
the LGAs mentioned in section 3.1. This process involved identifying treatment sites from the hard
copy information and entering this information into Arc View. In the process of identifying the
treated sites, the sites were also labelled as undergoing a specific speed zone change, for example
75km to 80km, for use in the analysis phase. As a result of processing the treatment site
information in Arc View, a list of location details in the form of Road Reference Points numbers
(RRPs) and Road Segment numbers (RSs) were obtained. Each RRP and RS was labelled with the
treatment site to which they related and the speed zone change type that occurred at that site.
4.2
CRASH DATA EXTRACTION
4.2.1
Treatment Crash Data
Having obtained the location details, the RRP and RS numbers were used to identify crashes at the
sites with speed zone changes, before and after implementation. This process involved merging the
RRP and RS numbers obtained from Arc View onto the Police recorded crash database. Crash
records whose RRP or RS numbers matched those of treated sites were then labelled as occurring at
specific treatment sites, including the type of speed zone change which occurred at each site.
Crashes whose RRP and RS numbers did not match were then labelled as suitable for use as control
crashes. The date of each accident within the treatment groups was then examined, labelling the
accident as before or after the speed zone change.
4.2.2
Control Crash Data
Having labelled those crashes occurring at speed zone change sites, the remaining crashes were
suitable for use as control crashes. As discussed, the reason for using control crashes is to adjust in
the analysis for the effects of other factors, apart from the speed zone changes, which have
influenced crash frequency in the study period. Hence the sites selected for controls should be as
similar as possible to the treatment sites apart from undergoing the change of interest (viz. a speed
zone change).
Treatment site descriptions provided by VicRoads provided very little basis for exact matching of
control sites. Consistent information provided for each treatment site which may have been useful
for control matching consisted only of the LGA in which the change occurred and the speed zone
before and after the change. Noting this, it was decided that control crashes would be matched on
two criteria; (1) LGA of treatment site, and (2) broad speed zoning before and after change. The
first criteria controls for the effects of other broad road safety programs in operation (such as speed
camera usage and Random Breath Testing) as well as economic effects in the local district. Criteria
Ev ALUA nON
OF THE CRASH EFFECTS
OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993
9
2 controls for specific features such as road type and adjacent land usage that are factors which
determine the allocated speed zone of a particular road. Given the range of speed zone changes
recorded at the treatment sites, this control matching strategy allowed all of the remaining crashes
not occurring at treated sites in each LGA to be used as controls which was useful in maximising
analysis statistical power.
Interrogation of the crash database showed the maj ority (approximately 90%) of pre 1993 crashes
occurred in either 60 or 100 km/h zones in both Melbourne and the rest of Victoria. Most of the
remaining crashes occurred in 75 km/h zones (which have, of course, changed) with only a few
occurring in 80 and 90 km/h zones. To ensure sufficient crash numbers in each control group, it is
essential that each control group selected included crashes in either 60 or 100 km/h zones. Given
this, Table 6 details the matching criteria used with relation to speed zone for each specific type of
speed zone change being evaluated.
TABLE 6 : Speed zone changes and proposed control matching speed zones of crashes in the same
LGA
speed
zones
(km/h)
within
75100
to
60
60
90
to
80
60
75
to
70
80
75
80
to
90
100
80
100
to100
90
Speed Zone 100 to 60,
80,90,100
80,90,100
to
70,
90,
60100
to60,80,90,100
50,in crash
Proposed
control
80,
90,
100
(+
60
Speed
Zone
Change
(km/h)
80 to 100,60
to 100
metropolitan
Melbourne)
o 60, 70 to 80,
sameLGA
The "other" category in Table 6 contains speed zone changes which have occurred at only 1 to 2
sites in the state. Given their small number it was not considered possible to evaluate these speed
zone change types individually to produce meaningful results, hence they were aggregated into a
residual category, labelled "other", for analysis. 100 km/h sites which had been rezoned to 110
km/h, whilst not the subject of this evaluation, were excluded as potential controls for this study.
10
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
All accidents had been labelled as occurring before or after the speed zone change (time) and in
either treatment or control groups (group). The file was then aggregated by the speed zone change
type, time and group within each of the LGAs selected to obtain a crash frequency for each
combination ready for analysis.
4.3
STATISTICAL ANALYSIS METHODS
The most widely used method of evaluating countermeasure effectiveness when control groups are
used for comparison, and one which would seem appropriate for use here, is that proposed by
Tanner (1958). Tanner's method has been used in many previous evaluations of accident black-spot
treatments, for example by Corben et al. (1990), Bui et al. (1991) and Tziotis (1993). In summary,
Tanner's method assumes that accidents at a treatment site are assigned to the before and after
treatment periods according to a Binomial distribution. Under the null hypothesis of no treatment
effect the ratio of accidents in the before to after period in the treatment group should be the same
as that in the control group. A chi-squared test for differences in the control versus treatment group
is then applied to test the null hypothesis of no treatment effect. Tanner's method includes a test of
equal treatment effect across a number of treated sites. The main advantage in using Tanner's test is
that it makes good use of information from control groups in the analysis.
A test similar in philosophy to that proposed by Tanner has been described by Bruhning and Emst
(1985). This test procedure uses the much newer and more elegant theory of Generalised Linear
Models, techniques that were not widely available at the time Tanner devised his method. This
new method also makes good use of information in the control group but is not restricted by the
assumption that the control group crash frequencies are error free as Tanner assumes. The method
of Bruhning and Emst recognises that the crash frequencies at the control sites will also be subject
to variation. A brief outline of the test proposed in Bruhning and Emst (1985) follows.
The accident data for a particular treatment site and control site in a simultaneous before and after
comparison can be summarised in a 2x2 contingency table, as in Figure 1.
FIGURE 1 : 2x2 crash data contingency table for a single treated site and its control
Before
n22
nl2
After
n2l
nll
where nij is the number of crashes in cell ij of the table. For L treatment sites, this may be
summarised in a series ofL 2x2 contingency tables as in Figure 2.
FIGURE 2 : 2x2 crash data contingency table for L treated sites and controls
Site
Ev ALUA nON
Before
After
nLll
nU2
nI22
n1l2
nl22
n22l
n2l2
nl2l
Before
n222
Control Group
Treatment Group
nUl
nlll
n211
OF THE CRASH EFFECTS OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993 11
A log-linear model with Poisson error structure, appropriate for the variability in the count data, of
the form
In( nijk) = fJo + fJi + fJij +fJik + fJijk
where i is the site number, j is the treatment or control group index and k is the before or after
treatment index, is then fitted to the data in Figure 2. Significant treatment effect in each group, i,
is then assessed by testing the significance of the fJijk parameter for each group. The magnitude of
the treatment effect is also assessed by the magnitude of this parameter and suitable confidence
limits can be calculated. Subtle modifications of the above model can be used to test for
homogeneity of treatment effect across a group of sites as well as each of the analysis levels
described in section 2.3. These modifications are described in Bruhning and Emst (1985).
4.4
POST HOC ANALYSIS
Having established any significant change in casualty crash rates associated with the implemented
speed zone changes, further post-hoc analysis of the data to determine the specific nature of the
changes was undertaken. The analysis presented is descriptive, examining differences in the
profiles of various crash descriptors before and after the speed zone change at the treated sites. The
key variable examined is DCA type, which describes the type of crash recorded in terms of vehicle
movement. The level of detail in the analysis was determined by the findings of the crash frequency
analysis.
In addition, results of the crash frequency analysis have been compared to the summary of speed
monitoring data at sites with changed speed zoning, detailed in Appendices F and 0 of the
Parliamentary Inquiry report (Road Safety Committee, 1995).
5.0
RESULTS
5.1
METROPOLITAN MELBOURNE
5.1.1
Crash Frequency
Results of the analysis of the effects of the speed zone rationalisation on casualty crash frequency
in metropolitan Melbourne are presented here. The analysis methods of Section 4 above were
employed to obtain the results.
Table 7 details the results of the casualty crash frequency analysis performed for metropolitan
Melbourne. Results are presented for the first three levels of analysis detailed in Section 2.3. That
is, the total program effect across all speed zone changes, general effects of speed zone increases
and speed zone decreases and the effects of each type of speed zone change across all LOAs of
metropolitan Melbourne. Although the amount of before and after crash data available was limited,
the fourth level of testing proposed in Section 2.3 which examines each type of speed change
within each LOA was carried out with results presented in Appendix C. Most of the estimates of
crash effects shown in Appendix C have relatively wide confidence limits, reflecting the small
quantities of data and prompting caution in their interpretation.
12
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
The estimated percentage change in casualty crash frequency for each hypothesis tested are shown
in Table 7 along with 95% confidence limits on each estimate. In interpreting these results, a
negative sign on the result indicates casualty crash frequency reduction whilst positive estimates
indicate crash frequency increase. As well as the 95% confidence limits and point estimates, the
significance level of the test of the null hypothesis is given. The significance level value is the
probability that the null hypothesis is true, viz. no change in crash frequency due to speed zone
changes, given the data tested. Small values of the significance level indicate significant casualty
crash frequency change due to speed re-zoning. For completeness, Table 7 also shows the number
of crashes in the combined before and after treatment study periods for both the treatment and
control sites. The number of cases in each of these cells gives an indication of the statistical power
available in each analysis, with more cases giving greater power to identify statistically significant
casualty crash frequency change.
TABLE 7 : Results of crash frequency analysis - Metropolitan Melbourne
MELBOURNE
Percentage
Change*
95% Confidence
Limits
Lower
Upper
Limit
Limit
Sig.
Level
75
2005
2210
0.3898
-18.62%
Control
Treatment
5174
7222
43
2031
-13.35%
2234
165
4122
432
6874
3359
1547
268
324
147
26
7295
4121
-0.13%
1433
22735
13230
0.2939
0.9834
0.3876
546
29.88%
12.60%
-13.40%
-11.42%
5.77%
-33.70%
-49.01%
-4.30%
13.25%
Number
of casualty
Speed
Changes
crashesZone
inoranalysis
hanges Across
nge
Increase
Decrease
All
90 Accidents
- 80change
Increase
Speed
Change control group crash trend
*:Net percentage
afterZone
adjusting/or
"*'"
.
Ev ALUA nON
OF THE CRASH EFFECTS
OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993
13
Examination of the results presented in Table 7 shows only two statistically significant hypothesis
test results at the 5% level of significance for speed changes in metropolitan Melbourne. Of the
individual speed zone changes, the changes from 100 to 80kmlh and 75 to 80 kmIh resulted in a
casualty crash change significant at the 5% level. For the 100 to 80km/h speed zone change, a
decrease in casualty crash frequency of 45.81 % was associated with the speed zone change, with a
null hypothesis significance probability of 0.0007. For the 75 to 80 kmIh speed zone change, an
increase in casualty crash frequency of 10.72% was observed with a significance probability of
0.0157.
Of the other individual speed zone changes analysed, one other showed a result worthy of note. The
decreased speed zoning from 75 to 60 kmIh was associated with an increase in casualty crash
frequency of 42.65% percent, with a null hypothesis significance probability of 0.0923. Whilst this
significance probability is not less than 5% it is sufficiently small to be described as marginally
statistically significant (significance probability between 0.05 and 0.10), indicating a likely effect
associated with this speed zone change. Whilst this result may look counter intuitive, there are
possible explanations for the observed crash increase, such as an increase in the variance of travel
speeds on the re-zoned road section, which will be further discussed below.
All the other individual speed zone changes with crash change results shown in Table 7 had null
hypothesis test significance probability greater than 0.10 indicating that no statistically significant
changes in crash frequency due to speed zone changes were found. As with all hypothesis testing of
this kind, a non significant hypothesis test result does not necessarily indicate there was no real
change in crash risk, but rather that a statistically significant change could not be identified given
the available data.
Table 7 also shows the results of testing the effects of speed zone changes in metropolitan
Melbourne, broadly classified by zones with increased speeds and zones with decreased speeds. For
roads where speed zones increased, an increase of casualty crash frequency by 9.3%, with a
statistical significance probability of 0.0076 was observed. For roads where speed zoning
decreased, no statistically significant change in casualty crash frequency was found (significance
probability = 0.3898).
The final result presented in Table 7 is an estimate of the overall effect of all speed zone changes on
casualty crashes in metropolitan Melbourne. The estimated net impact of all speed zone changes
analysed in metropolitan Melbourne, as shown in Table 7, is a 4.65% increase in casualty crash
frequency with a marginal statistical significance probability of 0.0969.
Using the proportion of sites sampled for analysis in Melbourne and the before treatment crash rate,
the estimated 4.65% increase in casualty crashes numbers across all Melbourne represents
estimated increase in the order of 235 casualty crashes per annum due to all speed zone changes.
Examining individual speed zone changes, an increase of 10.72% in casualty crash frequency
estimated for the 75 to 80 kmIh speed zone change represents in the order of 188 casualty crashes
per annum whilst an increase of 43% in casualty crash frequency estimated for the 75 to 60 km/h
speed zone change represents approximately 151 casualty crashes per annum across all Melbourne.
A decrease of 46% of casualty crash frequency for the 100 to 80 kmJh speed zone change translates
to a saving of approximately 44 casualty crashes per annum.
14
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
In summary, analysis of the effects of speed zone changes on casualty crash frequency in
metropolitan Melbourne led to the following results. Overall, speed zone changes were estimated to
increase casualty crash frequency by 4.65%, with a marginal statistical significance. No statistically
significant change in casualty crash frequency was found overall at sites where the zone speed
decreased, whilst a 9.25% casualty crash increase (with high statistical significance) was observed
at sites where zone speed increased. For particular speed zone changes, the change from 100 to 80
km/h showed a highly statistically significant casualty crash reduction of 47%, increased speed
zoning from 75 to 80 km/h showed a statistically significant casualty crash frequency increase of
10.7% and the change from 75 to 60 kmIh showed a marginally statistically significant casualty
crash increase of 43%.
5.1.2 Post Hoc Analyses
Having analysed the changes in casualty crash frequencies above, further analysis has been
undertaken in attempt to better understand these results. Further analysis for the metropolitan
Melbourne area has centred in two areas; (1) Examination of the changes in the profile of crash
types by Definition for Classifying Accident (DCA) grouping before and after speed zone change
for those sites with significant crash frequency changes, and (2) Relationship of the crash frequency
analysis results to results of speed monitoring reported in Appendices F and G of the
Parliamentary Inquiry report (Road Safety Committee, 1995).
Changes in Crash Type Profiles
Examination of changes in DCA patterns before and after speed zone changes for sites with
significant changes in casualty crash frequency has been carried out to determine if speed zone
changes have been responsible for changes in all crash types generally or only specific crash types.
DCA groups are specified by the first two digits of the DCA code. A full DCA chart is shown in
Appendix E. It should be noted that the analysis here looks only at changes in the distribution of
crashes by DCA at the sites where speed zones have statistically significant changes and makes no
reference to the control sites. To overcome the absence of reference to control groups, the analysis
examines changes in the proportion of crashes in each DCA group before and after treatment.
FIGURE 3 : Percentage of DCA groupings for crashes occurring in the treatment areas of
metropolitan Melbourne before and after speed zone changes: All speed zone changes.
IDBefore
_After
,
'"
~
8
~ ~
c
o
DCA
Ev ALUA nON
Group
;;
0Q.
'"0
c~
~~~~
'"
~
«
OF THE CRASH EFFECTS OF THE CHANGES
~
liS
c«
0~
~
Changes
Changes
E
IN SPEED ZONES IN VICTORIA
DURING
1992-1993
15
Figure 3 shows the distribution of crashes by DCA grouping for all speed zone change sites in
metropolitan Melbourne before and after speed zone changes. It can be seen from Figure 3 that the
distribution of DCAs has changed little as a result of the speed zone changes, with the proportion of
crashes in each group essentially the same before and after treatment. The analysis of crash
frequency change in Section 5.1.1 above estimated an overall casualty crash frequency increase of
4.65%. The DCA analysis in Figure 3 shows that it has been uniform across all crash types rather
than affecting any specific crash types.
Of the individual speed zone changes analysed for metropolitan Melbourne, only three showed
evidence of significant crash frequency changes. These were 100 to 80 km/h, 75 to 80 km/h and 75
to 60 km/h. Figures 4 5 and 6 show the results of DCA analysis for each of these three speed zone
changes respectively.
Figure 4 shows the results of DCA analysis for the 100 to 80 kmIh speed zone change where a
highly statistically significant crash frequency decrease of 45.81 % was observed. Figure 4 shows
notable decreases in the proportion of crashes of the types adjacent direction and opposing
manoeuvres (DCA 110-129), and relatively smaller decreases in the percentage of off-straight and
on-curve accident types (DCA 170-189). This suggests that the crash reduction attributed to this
speed zone change has been achieved by reducing crashes of these specific types. Whilst the
proportion of same direction manoeuvre crashes (DCA 130 -139) has increased markedly, it should
be recalled that the speed zone change resulted in an overall decrease in crash numbers.
Consequently, the proportionate increase is most likely not reflecting an increase in actual crash
numbers for that DCA type but rather reflects the decrease in the absolute crash numbers in other
categories.
FIGURE 4 : Percentage of DCA groupings for crashes occurring in the treatment areas of
metropolitan Melbourne before and after speed zone changes: 100-80km/h speed zone change.
8~-.5 ••
8.!!!.~
tu
RI
::J'S:
"
i;
.• E CLC::
15%
,-0.•
35%
20%
10%
o
::I
0 lc:(:e
'0
u
~tu8.8
~::J~
e•..
0u 25%
l..
.• ~.o
.•8
.5
J!l
'E g 5%
CL
••
Cl
c:: c::
c::
c::
CL
••
c::
o~
•• ::I
l/)
tu
RI ~
III Before Changes
30%
0%
• After Changes
; ;
::J
••
o
c::
RI
~~
••
o:;
.~
"" ::I
~ 8
~
DCA Group
16
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
c::
o~
:;
tu
a.
o
c::
~
o
::I
c::
o
.!l
c::
"
••
'H
«
For the change from 75 to 80 km/h, which resulted in an estimated 10.72% crash increase, Figure 5
shows small increases in casualty crashes of the types adjacent manoeuvres (DCA 110-119), other
manoeuvres (DCA 140-149), off-straight (DCA 170-179) and on curves (DCA 180-189), along
with a notable decreases in casualty crashes of the types manoeuvres involving pedestrians (DCA
100-109), crashes on path (DCA 160-169) and other crashes. Generally, however, the proportion of
crashes in each DCA type is similar before and after the speed zone change suggesting the change
has increased all crash types equally rather than any specific crash type.
FIGURE 5 : Percentage of DCA groupings for crashes occurring in the treatment areas of
metropolitan Melbourne before and after speed zone changes: 75-80km/h speed zone change
Cl Befors
_After
~
Changes
Changes
uc:~
Q.
c:
o
o
i1
..
~
«
DCA Group
FIGURE 6 : Percentage of DCA groupings for crashes occurring in the treatment areas of
metropolitan Melbourne before and after speed zone changes: 75-60km/h speed zone change
o.-'E >'"
20%
10%
30%
."
••
15%
0e~
::I
'E
> >e ••
'" c(
.5
'li
."
'8
«2lccl ~
~
:;:-~
e'"
!l0 C~~
51 5%
~~~
~
CL
CL
::I
.§
C»
l!l
;!
III Before Changes
0%
25%
.After
'"
Cl
.S;;;
'"
I!!
>
::I
e
o :;:'"
&.l!l
CL
e
.Q
'"
en
1d
.!::
:J
o•• e
l!l
E
"':;:IV
Cl)
s:::
iii
I!!
>
CL
::I
e
o
e'"
l!l
:;:
l;;
5
Il=
~ 1:
'E .2>
~~
SUi
«
••
~
o
::I
oe
Changes
l;;
s:::
CS
'"
'E
••
."
§
DCA Group
Ev ALUA nON
OF THE CRASH EFFECTS
OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993 17
Figure 6 shows the results of DCA analysis for the 75 to 60 km/h speed zone change where a
marginally statistically significant crash frequency increase of 42.65% was observed. Figure 6
shows increases in the proportion of crashes of the types opposing manoeuvres, same direction
manoeuvres, other manoeuvres and overtaking manoeuvres (DCA 120-159) with proportionate
crash decreases in the other DCA categories. Overall, the change in crash distribution across DCA
types before and after this speed zone change are small suggesting the effect on crashes is more
general than associated with any specific crash types.
Relation to Speed Monitoring
As part of the submissions made to the Parliamentary Inquiry into the revision of Victoria's speed
limits (Road Safety Committee, 1995), both the RACV and VicRoads submitted speed monitoring
data at a sample of sites where the speed limit had changed. Data was collected both before and
after the change. Appendices F and G in the Parliamentary Inquiry report detail the results of
analysis of the speed monitoring data. Whilst the data presented is representative of a limited
number of sites, it was considered useful to compare the published findings to the results of the
analysis undertaken here in an attempt to link any recorded speed changes with the estimated
changes in crash frequency.
Table 8 summarises the results of the speed monitoring data presented in the Parliamentary Inquiry
report along with the key results from the casualty crash frequency analysis described above. The
RACV state the sites monitored, with all being located in metropolitan Melbourne. The exact
location of the VicRoads sites, however, is not stated in the Parliamentary Inquiry report, hence it
was presumed these also lie in metropolitan Melbourne. Consequently, it was considered relevant
to compare all the speed monitoring results available to the results of the Melbourne casualty crash
analysis.
As shown in Table 8, the results ofVicRoads and RACV speed monitoring are generally consistent
in their conclusions. These results are also consistent with the results of the casualty crash
frequency analysis also detailed in Table 8.
For the speed zone change from 60 to 70 km/h, there was small mean speed increases observed on
undivided roads and no mean speed increases observed on divided roads. Whilst speed distributions
narrowed on the divided roads, the effects on undivided roads were less clear with some
distributions widening and others narrowing. No significant changes in casualty crash frequency
were observed for this speed zone change which is consistent with the results of speed monitoring
finding only slight increase or no mean speed change and generally an indication of a narrowing of
speed distributions.
Casualty crash analysis of the speed zone change from 75 to 80 km/h in metropolitan Melbourne
found a statistically significant 10.7% increase in crash frequency. Results of speed monitoring
suggest this may be due to the observed slight speed increases, particularly on undivided roads,
along with the a continued number of vehicles exceeding the speed limit excessively, particularly
on undivided roads, even at the new higher speed limit. In addition there is a suggestion of widened
speed distributions, again particularly on undivided roads, which is known to contribute to higher
crash risks.
18
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
From speed monitoring at sites with a speed zone change from 75 to 70 km/h, little change was
recorded in the mean travel speeds resulting in the reduced compliance with the new speed limit.
There was also little change in the distribution of speeds. These are both consistent with the
casualty crash frequency analysis results in finding no statistically significant change in casualty
crash frequency at sites with this speed zone change.
TABLE 8 : Summary of speed monitoring results presented in the Parliamentary Inquiry report
and related casualty crash frequency analysis results. Metropolitan Melbourne.
Increase
in
speeds
speeds
with
notable
Speeds
Slight decrease
increase
relatively
remain
unchanged
in
unchanged
speeds
at7780-85km/h
unchanged
RACV
VicRoads
RACV
VicRoads
10.72%
increase
%
increase
10.15
Half
and
half
sites
sites
narrowed
narrowed
widened
for
distribution
locations
ofwidening
8of
locations
6or
8
Distribution
in
out
8for
locations
Not
stated
Zone
Distribution
0.1045
significance
(-11.42,12.60)
(1.95,20.24)
0.0157
change
in
distribution
0.9834
Slight
inofno
Frequency
Casualty
Crash
Iinsignificance
Compliance
Speed
Mean
Speed
-0.13%
decreasee
and
half
sites
widened
r
of
excessive
it
but
VicRoads
ing
se.
limit
Greater
to
new
compliance.
eeding
limit
(-1.99,23.80)
mean
up
9km/h,
85th Analysis
less
drivers
below
limit
risk
drivers
the
same
m/h
gmained
toGreater
new
limit
and
limit
excessively
75-70
Results
75-80
Ev ALUAnON OFTHECRASHEFFECTSOFTHECHANGESIN SPEEDZONESINVICTORIADURING1992-1993
19
5.2
REST OF VICTORIA
5.2.1
Crash Frequency
Results of the casualty crash frequency analysis for speed zone changes in the rest of Victoria are
presented in Table 9. The format and interpretation of the results presented in Table 9 are the same
as for Table 7 above which details the results of the metropolitan Melbourne analysis. Results for
speed zone changes within individual LGAs in the rest of Victoria are given in Appendix D.
TABLE 9 : Results of crash frequency analysis - rest of Victoria
RURAL VICTORIA
95% Confidence
Limits
Percentage
Change*
538
344
Control
661
15.05%
12.40%
23.75%
658
1196
-11.80%
9.01%
0.8750
374
516
0.8110
0.6990
0.8122
-1.05%
0.6236
101.37%
99.21%
2.33%
-44.06%
-36.71%
-34.27%
13.20%
77.25%
24.41
6.42%
2117
0.8705
0.9710
0.5057
0.6572
0.6693
566
-1.89%
99.97%
27.67%
-25.56%
-15.39%
59.93%
-36.27%
70.28%
-35.92%
-39.07%
3657
4.17%
75.03%
-36.11%
-22.64%
Speed
Changes
crashes Zone
in analysis
75
70
All
Accidents
Increase Speed Zone Change
hanges Across
Lower
Limit
Upper
Limit
Sig.
Level
746
30
Treatment
15
119
185
159
22
Number
of casualty
*:Net percentage change after adjustingfor
control group crash trend
Whilst Table 9 shows the overall casualty crash frequency increase for all speed zone changes
combined in the rest of Victoria to be estimated as a 2.33%, the result is not statistically significant
leading to the conclusion that speed zone changes have not led to an overall change in crash
frequency in this region. Examination of each of the individual speed zone change types or
aggregations by increase or decrease speed zone change in Table 9 shows none of the speed zone
change categories considered to have resulted in a statistically significant change in casualty crash
frequency, supporting the overall result.
20
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
5.2.2 Post Hoc Analyses
Figure 7 details the results of DCA profile analysis for all speed zone changes combined in the rest
of Victoria before and after implementation of speed zone changes. Examination of Figure 7 shows
distributions of DCAs changed little as a result of speed zone changes with the proportion of
crashes in each group essentially the same before and after treatment. In conjunction with the lack
of statistically significant change in crash frequency observed in the rest of Victoria, the combined
results show that the speed zone changes in this region have not effected either crash frequency in
total or the distribution of crash types.
FIGURE 7 : Percentage of DCA groupings for crashes occurring in the treatment areas of the rest
of Victoria before and after speed zone changes: All speed zone changes.
o
.: >
~:;
cu'sm ::r
-0:
Q,C
'"
0:;
D- cc
o••
••
.•
'">c
'"-::r
.~
>~
30%
E __
O>~
~
U
'"
15%
&.8
t:
u 'tI
••
~
Q,
::r
0•..
.5
cl:
;;
20%
e..
00cl :;10%
lc
L
'" 8 lij c ~
5%
Q, C.!I
0"
08
en
fillBefore Changes
0%
25%
• After Changes
'"
~
>
::r
8c
••
:;
~
5
.<::
DCA Group
5.3
WHOLE OF VICTORIA
For completeness, the crash frequency analysis was also carried out for Victoria as a whole to
assess the casualty crash frequency effect overall. The results of the combined analysis are
presented in Table 10. Interpretation of the results in Table 10 is the same as Tables 7 and 9.
As expected from the relative number of crashes available for analysis in Tables 7 and 9, the
estimated casualty crash effects shown in Table 10 for the whole of Victoria are dominated by the
metropolitan Melbourne results. Hence the results of Table 10 closely follow those of Table 7 and
there were no different conclusions to those detailed for the analysis of speed zone changes in
metropolitan Melbourne.
Ev ALUA nON
OF THE CRASH EFFECTS
OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993 21
TABLE 10 : Results of crash frequency analysis - All of Victoria
VICTORIA
Percentage
Change*
2554
2031
165
7440
0.2939
13.25%
Control
8491
-13.35%
661
2608
4638
4779
970
546
1.24%
0.8349
13.67%
-33.70%
26392
15347
-9.84%
Changes
Speed
crashesZone
in analysis
All
Accidents
90
-80
Increase Speed Zone Change
Changes
Across
nge
95% Confidence
Limits
Lower
Upper
Limit
Limit
2027
3478
1577
324
272
26
75
6
5333
7407
47
Treatment
1448
154
Number
of casualty
*:Net percentage change after a4justingfor
6.0
Sig.
Level
control group crash trend
DISCUSSION
Selection of sites where speed zone changes had taken place for use in the analysis presented here
was largely dictated by the quality of the data describing the sites supplied by the VicRoads
regional offices. Of those sites with suitable data, the majority have been included in the analysis.
Ideally, when sampling sites for analysis, a completely random sampling mechanism is necessary
from a statistical point of view in order to eliminate any possibility of systematic bias in the data
affecting the results of analysis. As it was not possible to apply a completely random sampling
procedure for site selection, it has been assumed that the sites analysed are generally representative
of all sites of the same type and hence the analysis is representative of the true effects of speed zone
changes in Victoria.
Overall, the results of the evaluation presented here give indications of the effects that the revision
of Victoria's speed zone limits have had on casualty crash frequency. Results of analysis in
22
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
metropolitan Melbourne indicate that the increased speed zoning, with no co-incidental road
geometry changes, was generally associated with an increase in casualty crash frequency, reflected
particularly in the 75 to 80 km/h speed zone change. Speed monitoring results suggest this may be
a result of higher mean speeds, plus a wider distribution of speeds in some ofthese zones.
The results also indicate that decreasing speed zoning in metropolitan Melbourne did not generally
result in statistically significant change in casualty crash frequency. There are however some
notable exceptions when examining particular speed zone decreases. For the 100 to 80 km/h speed
zone change, a significant crash reduction of 45.8% was estimated. This result could perhaps
suggest that the original 100km/h speed zoning was too high on the roads concerned, with 80km/h
representing a travel speed with which more people are willing to comply. It should also be noted
that this speed zone change typically occurred in urban fringe areas on roads possibly not of a high
engineering standard. Speed monitoring data, which was not available for the study, would have
been necessary to support this hypothesis. The other notable speed zone decrease is the change
from 75 to 60 km/h zoning that appears to have increased casualty crash frequency. This is possibly
the result of the speed zone change increasing the variance in travel speeds with some drivers
refusing to reduce travel speed from that which they have been used to travelling at in the past.
Again, speed monitoring data would be needed to support this theory.
Assuming the estimated crash effects of speed zone changes in Melbourne for the sites examined in
this study are generally representative of the crash effects speed zone changes would have at any
site, the results point to some important implications for road authorities to consider for speed
zoning in the future. Firstly, casualty crash increases may be likely when increasing zoned speeds,
even by as little as 5km/h. Conversely, when reducing posted speed limits, there appears to be a
reluctance for drivers to uniformly reduce travel speeds to comply with the new limits, hence
leading to no change in crash frequency. In some cases, lowering of speed limits may lead to an
increase in the distribution of travel speeds consequently causing an increase in crash rates. A
means of lowering speed limits whilst achieving a corresponding reduction in crash frequency
through general driver compliance with the new posted limit may be to use speed enforcement,
such as speed camera or laser deployment, for a period of time following re-zoning.
Results of the analysis of crash effects of the speed zone changes in metropolitan Melbourne from
this study are largely consistent within statistical accuracy with the results of the earlier study by
Newstead and Mullan (1996) apart from two notable exceptions. This first is the change from 60 to
80 km/h which the original study estimated to produce a statistically significant crash reduction of
47% but here showed no statistically significant crash change. Reasons for this inconsistency are
not obvious. Sites with this speed zone change are typically sections of road around intersections
on major arterial roads which have been re-zoned to make them consistent with the rest of the road.
The change in result suggests there might be an initial crash effect straight after the speed zone
change which diminishes with time, a point further discussed in the context of the results for the
rest of Victoria below. Another reason may be changes in signal phasing at the intersections
concerned some time after the speed zone changes. Further research would be needed to establish
the exact cause. The second apparent inconsistency is the zone change from 75 to 60 km/h which
here was estimated to produce a marginally significant 43% crash increase whereas in the previous
study was estimated to produce no significant crash change. It should be noted for this result,
however, that the current estimate lies within the confidence limits of the previous estimate which
were very wide due to limited data. Hence the results are not inconsistent in a statistical sense.
Ev ALUA nON
OF THE CRASH EFFECTS OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993
23
Results of the present analysis for the rest of Victoria indicate that the speed zone revisions have
resulted in no statistically significant changes in casualty crash frequency either overall or for any
particular speed zone change. As noted, the speed revisions in this area generally have been
implemented in transition zones between 100 km/h open roads and 60 km/h rural towns with the
stretches of road effected being typically quite short. This has resulted in a relatively small number
of crashes available for analysis and has perhaps limited the potential for large crash effects. It is,
however, interesting to note that the magnitude of the crash effects estimated for the rest of Victoria
are, although not statistically significant, generally consistent with those estimated for metropolitan
Melbourne.
The results ofthe earlier analysis of the crash effects of Victoria's speed zone changes by Newstead
and Mullan (1996) reported an overall crash reduction of 32.9% in the rest of Victoria with
marginal statistical significance, which is apparently quite different to the results obtained here.
Estimated crash changes for individual speed zone changes also appear to be quite different
between this study and the previous one, although none of these results achieved statistical
significance in either study meaning there is no inconsistency in the conclusions drawn. There are a
number of possible reasons for the inconsistency in the estimated changes between the two studies.
The most likely of these is statistical estimation variance, highlighted by noting the very small
number of crashes at speed zone change sites in the rest of Victoria in Newstead and Mullan (1996)
(78 crashes in total) and also in the current study (185 crashes in total).
Another possible reason for the difference between crash change estimates in the rest of Victoria is
a change in compliance with new speed limits over time. It is possible drivers may comply well
with speed zone changes for a period immediately after implementation in response to the new
situation with compliance diminishing as time elapses and driver behaviour regresses to that
observed pre-change. This hypothesis is consistent with the large reductions estimated in the early
study, which focuses on the first year after speed zone changes, and the lack of change observed in
this study which uses data for a period three years after speed zone changes. This hypothesis could
only be tested if longitudinal speed monitoring data were available after implementation of the
speed zone changes.
7.0
CONCLUSIONS
Analysis of the effects of speed zone changes on casualty crash frequency in metropolitan
Melbourne showed an overall increase in casualty crash frequency of 4.7%, although this result was
of marginal statistical significance and should be interpreted with caution. Assessment of the
general effects on casualty crash frequency of increasing zoned speed showed a highly statistically
significant increase in casualty crash frequency of 9.3% but decreasing zoned speed showed no
statistically significant change in casualty crash frequency. For particular speed zone changes, the
change from 100 to 80 km/h showed a highly statistically significant casualty crash reduction of
46%. Increased speed zoning from 75 to 80 km/h showed a statistically significant casualty crash
frequency increase of 10.7% and the change from 75 to 60 km/h showed a marginally statistically
significant casualty crash frequency increase of 43%.
The results of analysis of casualty crash frequency in metropolitan Melbourne were generally
consistent with the results of speed monitoring.
24
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
Most of the speed zone changes which occurred in the rest of Victoria took place on the fringes of
country towns in the speed transition zones between 100 km/h zones of the open highway and 60
km/h zones of the built up town areas. Analysis of the overall casualty crash frequency change for
all speed zone changes combined in the rest of Victoria as well as analysis by specific speed zone
changes showed no statistically significant changes in crash frequency.
8.0
RECOMMENDATIONS
From the results of the analysis presented in this report, two recommendations are made.
1.
Continued monitoring of the effects of speed zone changes on casualty crash frequency
incorporating more rigorous monitoring of associated changes in traffic speeds
Results of this study have pointed to the general effects of speed zone changes on casualty crash
frequency. The range of specific speed zone changes for which definitive crash effect estimates are
available is, however, relatively small and the general relevance of the results to speed changes
outside the sites studied is unclear. Speed monitoring results from the Parliamentary Inquiry report
available for use in this study, whilst of some value in interpreting the estimated crash effects, were
not extensive enough to provide definitive insight into the mechanisms underlying the observed
crash changes.
It is recommended that the crash effects of any further speed zone changes carried out in Victoria in
the future be monitored using the approaches developed here. Opportunities to carry out extended
study in this area should arise considering there is continued re-zoning of road speed limits being
carried out, especially on city fringes where urban development is under way. It is also
recommended that extensive speed monitoring be carried out at sites where speed zones have
changed in order to better understand the mechanisms of any observed crash effects. In particular,
speed monitoring needs to be carried out longitudinally after speed zone changes to determine if
driver speed behaviour changes over time in response to the changes, as has been suggested by the
results of this study.
2.
Standardisation and review of data collection and reporting procedures used by VicRoads
regional offices.
A particular problem for the evaluation presented here was the lack of detail of some data supplied
describing the speed zone changes which were carried out. This limited the number of sites with
speed zone changes which could be included in the analysis, and hence the limited statistical power
of the analysis.
It is recommended that the issue of data collection methods be addressed for any future work on
this subject that may be carried out and for other prospective evaluation studies of this type. For
prospective studies it may involve the design and implementation of standard data forms and data
collection procedures across all regions from which data is supplied.
Ev ALUA nON
OF THE CRASH EFFECTS OF THE CHANGES
IN SPEED ZONES IN VICTORIA
DURING
1992-1993
25
9.0
REFERENCES
BRUHNING, E. & ERNST, G. 1985, 'Log-linear models in effectiveness studies - an application
to simultaneous before-after comparisons with control group', International Meeting on the
Evaluation of Local traffic Safety Measures.
BTCE 1993, 'Cost effectiveness of black spot treatments - a pilot study.' Bureau of Transport and
Communications Economics, Working paper No. 9, October 1993.
BUI, B., CAMERON, M. and FOONG, C.W. 1991, 'Effect of right turn phases at signalised
intersections Part 1 - Safety performance.', Report No. 20, Monash University Accident Research
Centre, Clayton, Victoria.
CORBEN, B.F., AMBROSE, C. and FOONG, C.W. 1990, 'Evaluation of accident black spot
treatments', Report No. 11, Monash University Accident Research Centre, Clayton, Victoria.
GREEN, P. 1991, 'Victorian Road Accident Database - Codebook for 1983 through 1991',
VicRoads Road Information Services, Accident Information Centre.
NEWSTEAD, S., CAMERON, M., GANTZER, S. and VULCAN, P. (1995) "Modelling of some
major factors influencing road trauma trends in Victoria 1989-93" Report No. 74, Monash
University Accident Research Centre.
NEWSTEAD, S. and MULLAN, N. 1996, 'Evaluation ofthe crash effects of the changes in speed
zones in Victoria during 1993-1994 (excluding 100 to 110 km/h)', Report No. 98, Monash
University Accident Research Centre, Clayton, Victoria.
ROAD SAFETY COMMITTEE 1995 'Inquiry into the Revision of Speed Limits' Parliament Of
Victoria, Victorian Government Printer, April 1995.
TANNER, J.C. (1958), 'A problem in the combination of accident frequencies', Biometrika, 45,
pp 331-42.
TZIOTIS, M. 1993. 'Evaluation of mid block accident blackspot treatments', Report No. 48,
Monash University Accident Research Centre, Clayton, Victoria.
26
MONASH
UNIVERSITY
ACCIDENT
RESEARCH
CENTRE
APPENDIX A
Sample of speed zone change sites used in analysis: Metropolitan Melbourne
METROPOLITAN
Sample of Speed Zone Change Sites Used in Analysis
~
p~l~
pv ~')
\--
~
-H
fL- _
~}:':
I
I
/-.
~
~I
=
~"-n'
J,.,
~)
/\~rK\
J
~
iJIn=:~
_
~
<
~/-r
L
-
-~
~
•
'"/11
I
';:j
"/
"" '-J7~~~~'
"
-,y
I"
0
~"'" '-~
~
F~'""y,~
~"'"'I
n (jj~ ~~
,~,J
"ij'
".
flr~~"
vt:.i~~_
1
f.fIU.
~
~-,_J ~"
T
~
I
J~
__
r
t
~j
f'
~~\J~
1\~ f-';
~up>;"d'~
-h
~~ ~~~y~~
rrttf
r
ru"
-c>"
r
"
/\I
li,~
Roads
'~.
t
(::.
'-
"-uS
--:;1':'1
-\ft-
~~[
:
rr I"~I../-
~~')re~ ljl
f;S,'
L>--l.\
~II
~t,:-r;Nt~-l~~1
fiM:
\
\....,.
r ~ ~<
<Z
r-1J-')~~6;~~
~ ~h~\l~
,~
..L
>'-
{
7i•.
.. t:ttf.;c¥'~
'I ~ 1''''' >
CJ Old LGA's
);~?".,
T" '1-
I
/ 'o/J 'PJ~>)-,J
.I.f'
~~ J
~t j
,}
Zone Changes
" -5 :-\ ~
'10,;.
_ 'J.. ~
~"
III
I
1 T'
Cc
,.J'r "
-y,
~,
j~
~
fll
,,~,
>
LEGEND
~ ~
~~1 •...,.,
~ \ ~.!lY;::'~
/1/-·11
~
~
~
-c
.~~~
I~
-~}tT7,i
:/
_I
~~
/
C
--f
"-
I, I'
1ff§
r
-if
-.....
MONASH UNIVERSITY
_I;:~
.-'\~,
'jl'''='~
rf
/-\
,,'~
Y':t1
+
I
Accident Research Centre
APPENDIXB
Sample of speed zone change sites used in analysis:
Rest of Victoria
RURAL VICTORIA
Local Government Areas Used in Analysis
+
MONASH UNIVERSITY
Accident Research Centre
APPENDIXC
Listing of speed zone change sites evaluated in metropolitan Melbourne, length
of road covered by each speed zone change and estimated crash effects by speed
zone change type within LGAs
MELBOURNE
"""'"",l0"".,., ...,.,.,
______
, ..,"""','>~
75 to 80
60 to 60
75
70
"".',',-"'"
., •.'.'.'.'.-.,-__________
',.'"
./-,-._._.
-...
-
1ii'.5~II.ii.18.~
242.18%
0.9702
217.66%
-68.14%
-38.23%
97.79%
0.5317
0.5905
0.3615
-60.87%
-69.40%
-93.84%
0.9783
4.00%
1347.56%
1083.19%
650.41%
262.73%
-67.21%
-51.89%
-95.86%
-69.95%
0.5729
0.2638
0.8047
3258.47%
0.3368
56.86%
60
142.56%
-73.61%
0.6933
0.8211
-27.66%
-20.00%
1354.40%
1492.93%
1519.90%
-94.25%
384.07%
533.03%
-22.16%
-47.78%
77.04%
63.44%
-51.11%
-57.17%
-8.45%
-45.17%
0.5129
0.4384
0.1145
0.1548
0.3476
0.9932
0.9805
0.9373 2354.92%
5088.20%
0.1492
433.35%
-21.36%
-24.90%
-10.61%
44.04%
94.12%
-1.20%
-3.45%
-37.19%
-29.68%
133.11%
-90.77%
-95.90%
73.50%
-69.12%
-39.78%
-89.11%
18.04%
0.3202
0.8682
0.4180
0.4656
0.8508
2446.38%
0.7656
-57.97%
75
-30.55%
20.00%
53.33%
17.94%
22.44%
-66.59%
3566.64%
0.2959
-36.68%
-45.10%
-30.20%
180.65%
45.12%
40.46%
48.13%
60.89%
-60.08%
65.18%
-65.96%
71.72%
-63.70%
0.8693
0.8617
0.8086
0.8089
0.8997
-27.42%
-28.99%
-0.99%
-9.35%
-4.77%
4.28%
1485.45%
0.3528
0.9849
25.96%
-2.66%
1427.23%
-31.63%
348.85%
-94.47%
47.35%
-96.59%
19.94%
0.3278
0.9851
0.4510
0.9530
1.0000
-8.11%
0.37%
1654.79%
2.33%
-47.82%
-76.23%
0.7241
0.9151
-30.00%
163.90%
-9.45%
-7.14%
1097.62%
-95.63%
126.61%
1129.46%
-93.87%
44.38%
-95.58%
-94.50%
0.7156
0.8316
-10.59%
-26.32%
81.81%
522.47%
331.32%
104.93%
67.83%
-71.26%
0.6355
-20.45%
-15.69%
-17.65%
4.39%
61.90%
-55.99%
0.6099
-15.59%
-90.32%
-84.45%
0.1152
2474.98%
0.7484
250.01%
-56.36%
57.90%
90
80 LGA
-59.06%
-55.31%
100.69%
-90.79%
-92.95%
0.5336
0.3646
0.9559
2469.61%
0.7643
-19.86%
12.00%
20.00%
-94.02%
-25.53%
104.96%
-22.59%
71.51%
-70.63%
91.85%
0.4462
0.4599
-29.03%
19.53%
to 70
-95.93%
-41.97%
-16.57%
0.00%
57.12%
1943.09%
Speed Zone
Changes
Within
Each
;,53.85%
60 to 80
TEMPLESTOWE
BERWICK
Percentage
Change*
Lower
Limit
Upper
Limit
Pr> Ch-Sq
MELBOURNE
Speed Zone Changes
KNOX
MELBOURNE
PRESTON
RICHMOND
SUNSHINE
Percentage
Change*
Lower
Limit
60 to 70
5.90%
-39.01%
83.87%
0.8387
75 to 80
15.52%
-26.27%
8fOO%
0.5289
0.9216
Within
Pr > Ch-Sq
Each LGA
Other
4.83%
-59.03%
168.27%
75 to 70
23.59%
-15.57%
80.91%
0.2761
60 to 70
47.12%
-18.51%
165.61%
0.2001
75 to 80
-2.87%
-23.54%
23.39%
0.8118
Other
-56.91%
-88.10%
56.04%
0.1997
75'Tci'70
60 to 70
75 to 80
75 to 70
5.96%
-42.42%
95.00%
0.8524
-8.01%
-42.93%
48.28%
0.7317
41.79%
-8.02%
118.58%
0.1137
-9.32%
-44.06%
47.01%
0.6915
60 to 70
60.32%
.43.8§%
235.33%
41.13%
0.2100
75 to 80
-23.35%
-77.69%
60 to 70
147.76%
-50.19%
1132.41%
0.2677
75 to 80
18.44%
-23.83%
84.15%
0.4524
'751070
60 to 70
10.49%
-84.64%
694.65%
0.9211
18.54%
-21.81%
79.72%
0.4231
60 to 80
-18.75%
-74.98%
163.89%
75 to 60
125.96%
-77.09%
2128.45%
75
WAVERLEY
Upper
Limit
to
0.2195
80
Other
-33.68%
-84.67%
186.91%
75 to 70
20.29%
-18.87%
78.33%
60 to 70
20.00%
-21.45%
83.32%
751080
::ro.62%
-36.29%
25.38%
60 to 80
75 to 60
-36.78%
-79.24%
92.53%
*;
Net percentage change after adjusting for control group crash trend
••••
~Mlli~
MELBOURNE: LENGTH OF ROAD BY $PEl:D ZONE CHANGE IN EACH LGA
LGA NAME
BERWICK
BRIGHTON
Other
75
to
60
60
Other
70
80
100
60
to
to60
70
90
100
75
60
80
100
to
80
60
to
80
75
60
90
to 70
80
90
80
60to
70
BULLA
CAMBERWELL
CHELSEA
COLLlNGWOOD
CRANBOURNE
CROYDON
DANDENONG
DIAMOND VALLEY
SPEED ZONE CHANGE
90 to 80
100 to 80
60 to 70
75 to 80
60 to 80
100 to 90
75 to 60
Other
TOTAL ROAD LENGTH (metres)
11312
7493
1232
40834
1350
2794
1250
1062
60 to 70
387
530
1132
2403
615
20882
695
201
14550
3514
3925
1019
860
978
4427
2470
2297
1289
2245
7451
1663
599
6138
658
978
31607
11991
11976
4818
6316
5875
8141
1020
730
225
988
1169
0to
80
Other
0Other
to
to
60
80
70
70
5
toto
60
80
70
MELBOURNE:
LENGTH
ROAD
BY
ZONE(metres)
CHANGE
792
1912
49016
832
4449
7401
564
964
75
7282
6835
to
70
1124
34821
18135
2674
2824
2377
1255
1116
1923
1978
1963
80 SPEED
2077
8070
8577
5471
9121
18686
12266
410
868
14161
16202
3470
5750
1406
SPEED OF
ZONE
CHANGE
TOTAL
ROAD
LENGTH
TEMPLESTOWE
LGA NAME
IN EACH LGA
MELBOURNE: TOTAL ROAD LENGTH7206
BY SPEED ZONE CHANGE
97566
70164
47390
224810
17546
11251
9302
5626
TOTAL
ROAD LENGTH (metres)
Other
100
75
60to
to
to60
70
90
80
75
60
to
80
70
90 to 80
SPEED ZONE CHANGE
MELBOURNE:
SPEED ZONE CHANGEA'BECKETT
SITES
USED
IN
EVALUATION
3312
988
743
476
273
745
264
1631
320
1049
419
1856
1277
759
1882
HARKAWAYRD
1232
444
4835
2794
468
1262
722
482
2175
1444
1406
1399
959
2395
HALLAM
RD
580
3095
5258
1259
2831
1350
1644
1482
1519
702
1449
1500
1502
956
1150
RIDDELLRD
530
4278
RIDDELL
RD
1019
610
5670
978
HEATHERTON
NARRE
WARREN
RD
NORTH
CASERTADVE
RD
JACQUES
INGLlS
PARKHILL
RD
RD
DVE
201
MANUKARD
MEMORIAL
BAY
ST
HWY
DVE
EAST
2895
1026
KINGRD
HALLAMRD
CRANBOURNE
FULLARD
GOLF
ERNST
LINKS
WANKE
RD
RD
-RD
NARRE
WARREN
HARKAWAYRD
CHURCHILL
PARK
RD
RD
CENTRERD
CENTRE
WEBBST
2064
BELGRAVEHALLAM
RD
RD
LENGTH
ROAD
ROAD
NAME
RD
ARDLlE
OLD
RIDDELL
ST
RD
MICKLEHAM
HILLVIEWRD
LOEMANSRD
BULLARD
CRAIGIEBURN
RD
WEST
POUNDRD
BELGRAVE
KINGRD
MULGRAVE
L1GHTWOOD
HALLAM
HEA
BOURNEVALE
BAKERRD
100to
THERTON
HALLAM
FWY
DVE
RD
BULLA-DIGGERS
REST
RD
SOUTHRD
BULLARD
WARRAWONG
KIRKHAMCRT
WARREN
RD
NEPEANHWY
NARREINGLlS
O'GRADYRD
RD
NORTH
HARSWOODRD
ROBINSONRD
KINGRD
PRINCESS
SHRIVES
RD
CENTRE
COUNTY
LANCASTER
DEEWYRD
POUNDRD
Other
RD
DVE
AVE
POWERRD
POUNDRD
RD
ASHFIELD
DVE
CRANBOURNENARRE
WARREN
RD
ENTERPRISE
CENTRE
RD
AVE
MANUKARD
BERWICK
BEACONSFIELD
CANISTON
MANUKARD
RD
AVE
MAINST
CHANGE
PHILLlP
DVE
ARDLlE
ST
DORCHESTER
ST
CALDERFWY
PRINCESS
CASERTADVE
CRINNIONRD
FWY/HWY
EAST
GLENDORAAVE
HARTLEYST
AVONDALE
L1GHTWOOD
HESSELRD
60
to
70
70
ST
CRT
BULLAQUARTZST
DIGGERS
REST
RD
GLENHUNTL
GLENDORAAVE
to
90
Y
RD
HEATHERTON
LOGON
ROBINSONRD
PARK
RD
MEMORIAL
BEACH
EBELI
MOULEAVE
CLOSE
RD/ST.KILDA
DVE
ST
VESPER
PROGRESSST
SOUTH
VESPER
GIPPSLAND
DVE
HWY
HILLSMEADE
HALLAM
100
80
NORTH
DVE
RD
60
60
HOMESTEAD
RD
WEST
to80
OF
WEBB
ST
PRINCESS
MILLWOODAVE
CANISTON
MILLWOOD
HWY
AVE
AVE
EAST
GLENWOODRD
75
KENNEDYAVE
GLENWOODRD
ROAD
INTERSECTION
1RAMP
RIDDEL
RD/DAL
YRMPLE
RD
FAWKNERST
SOUTHAVE
BELGRAVEBEACONSFIELD
HALLAM
-(NORTH
EMERALD
RD
RD
BELLGRAVE
-RD
HALLAM
NARRE
RD
CRANBOURNEWARREN
RD OF
NARREWARREN
NORTH
90
to
ROAD
INTERSECTION
2
CIMBERWOOD
DVE
(HALLAM
NORTH
RD)
CRANBOURNE
RD
(CENTRE
RD)
(metres)
CLOVER
CLOSE
OF
...
)...)
PROGRESS
(EAST
OF
...)
COUNTY
DVE
(0.5KM
WEST
OF
..)
KIRKWOOD
CRESENT
(SOUTH
EAST
...)
COUNTY
HAGEN
FULLARD
DVE
DVE
RD
(25M
(CENTRE
(0.5KM
WEST
WEST
RD)
OF
...)
HALLAM
RD
(0.28KM
EAST
OF
...)
BULLA
RD
(1.05KM
OF
...)
(0.92KM
WEST
OF
...)
BERWlCKCRANBOURNE
RD
(CLYDE
RD)
BEACONS
FIELDEMERALD
RD
(NORTH
WEST OF ...)
ZONE
SPEED
MELBOURNE:
SPEED ZONE CHANGE THE
SITES
USED
IN
EVALUATION
4B2
756
311
19B6
364
97B
1449
2394
64B
WARRIGAL
HWY
FAKENHAM
RD
14B9
6Bl
CRAIGIEBURN
RD
2415
1239
445
940
VINEYARDRD
OLD
CALDER
HWY
3B7
519
EASTERNFWY
HEIDELBERG
RD
BULLEEN
RD
557
713
1266
BALLARTO
RD
1023
1021
670
1273
B60
133
4BO
773
GAP
RD
696
1119
1135
607
441
625
2352
B52
l09B
6B6
706
657
HUMEHWY
TRANSPORT
MOUNT
PUNJEL
COOPERST
RIDLEY
DVE
DVE
RD
WlLLWOODRD
EASTERN
DONCASTER
FWY
RD
RAMP
THOMPSONS
FERDINAND
DONCASTER
BO
AVE
NEPEANHWY
ROXBURGH
PARUNAST
McDOUGALL
PROGRESS
RAILWAY
75
to
60
PARADE
AVE
PARK
RD
DVE
1030
SUNBURY
BAXTER
TOORADIN
RD
RD
FRANC
REGENTST
100toBO
60
IS
BOULEVARD
THEGAPRD
ABBOTTS
SUNBURYRD
BROWNSRD
HO
DD
LE
ST
RD
EAST
COLLEYST
PARK
OF
DVE
POWLETT
ST
LENGTH
ROAD
376
RIDDELLRD
ROAD
NAME
1400
CENTRE
BERWICK
RD
CRANBOURNE
RD
BO
CHANGE
HOMESTEAD
RD
3B79
RD
EVANSRD
HALLRD
HALL
RD
NARRE
WARREN
VICTORIARD
DANDENONG
RD
VALLEY
HWY
331
75toBO
2340
POUNDRD
B39
McDOWALLRD
RD
930
HALLAM
POTTS
KIMBERLEY
DVE
1196
16B5
VALLEYDVE
20Bl
1294
2110
2120
357
SEARCH
BLACKWELLS
GREYTHORN
HIGH
BULLARD
DIGGERS
60
ST
REST
RD
RD
-Y
COIMADAI
RD
TAYLORAVE
HUMEHWY
MOORERD
70
TURNBULLST
ARDCLONEY
THE
CALDERHWY
ESPLANADE
DVE
SPAVINDVE
HEALESVILLE
PEARCEDALE
EASTERN
WlLSONS
VICTORIA
LANGLEY
Other
BO
LANE
RD
FWY
PARADE
-EAST
RAMP
KOO
WEE
RUP
RD
SOMERTON
RD
FRENCH
BUNDS
CRANBOURNENORTHRD
SOUTH
90
LANE
GATEWA
RD
NARRE
WARREN
RD
SOMERTONRD
CRANBOURNE
CANNONS
CREEK
-FRANKSTON
FRANKSTON
RD
BAXTERCURRAWONG
ROAD
to
RD
INTERSECTION
TOORADIN
GROVE
RD
2
McCLELLAND
100
to
DANDENONG
HASTINGS
RD
SOUTH
CAMPRD
GIPPSLAND
HWY
lOOtoBO
HALLRD
McCORMICKS
RD
MUNDAYCRT
HERITAGE
75
60
DVE
CENTRALRD
RD
POUNDRD
FWY
RAMP
POUND
RD/SHRIVES
RD
TARELLAST
CODRINGTON
LEXTON
DVE
ST
SOUTH
GIPPSLAND
HWY
LESDONAVE
KINGSWOOD
DVE
ALDRIDGE
.93KM
CAHILLST
MAGNOLIA
NORTH
BOULEVARD
CAHILL
ST
ROAD
MICKLEHAM
MIDDLERD
INTERSECTION
RD
1WEST
CENTRE
SCOTTST
RANGEVIEW
BRINNINGS
KIMBERLEY
HALLRD
DVE
O'GRADY
RD/CENTRE
RD
CAIRNSRD
POTTS
GRAIN
STORE
CRT
QUARRYRD
CLARENDON
ST
TARELLAST
SUNBURY
RD
(MACEDON
ST)
(FRANKSTON-DANDENONG
RD)
(metres)
McDOUGALL
CRAIGIEBURN
RD
RD
(SOUTH
(0.5BKM
OF
NORTH
...)
OF
...)
SUNBURY
RD
(HORNE
ST)
KOO
WEE
RUP
RD
(1.25KM
WEST
OF
HEALESVILLE
EAST
NARRE
ST
WARREN
(0.15KM
(0.46KM
EAST
RD
(1KM
OF
OF
...)
EAST
...)
OF
...)
ELANDRA
WAY
(0.4KM
NORTH
OF.)
EVANS
(0.2KM
WEST
OF
...)
OLD
CALDER
HWY
BULLADIGGERS
REST
RD)
POWLETT
ST
(EAST
OF...)
(10M
EAST
OF.)
WETHERSDANE
DVE
(0.25KM
EAST
OF.)
CAMERON
RD
(CRANBOURNE
NARRE-WARREN
BERWlCKCRANBOURNE
RD
(0.2KM
SOUTH
OF.)RD)- K-W-RU
ZONE
60 to 70
SPEED
MELBOURNE:
SPEED ZONE CHANGE THOMPSONS
SITES
USED
IN
EVALUATION
2520
1596
454
1169
1325
2417
456
2977
3311
766
600
325
677
632
646
409
495
313
329
2443
1382
1080
ROAD
1465
1757
1289
944
584
2587
1082
695
GRANTS
RD
CANTEBURY
BALLARTO
BAYSWA
TER
RD
RD
COVEN
AVE
2755
1475
DORSET
DORSETRD
DORSETRD
4400
3874
1554
751060
BAYSWATER
RD
1359
844
730
HEATHERTON
1294
204
563
276
390
350
1321
642
510
BRD
GULLY
RD
1635
SOUTH
WARRANDYTE
CENTRE
GIPPSLAND
RD
RD
FWY
RIMFIRE
PRINCESS
BEUNALRD
DVE
HWY
EAST
LENGTH
WARNEETRD
ROAD
NAME
ARUMAST
GENTLE
ANNIE
DVE
ALDERST
751080
RD
CHANGE
HWY
GOLF
McCLELLAND
NORTHRD
LINKS
RD
DVE
BARRETS
BALLARTRD
McCLELLAND
ALDERSHOT
1001090
DVE
DVE
TAYLORS
COLCHESTER
MAROONDAH
HWY
WILBELMINA
CRT
CHELTENHAM
GREAVES
BENNETST
CHANDLER
ST
RD
MOUNT
DANDENONG
CANTEBURY
CLEGGAVE
GREENSRD
COLCHESTER
MOUNT
ORDISHRD
DANDENONG
RD
GREENS
HAMMONDRD
CAHILLST
601080
1020
407
434
862
PRINCESS
MULGRAVE
HWY
FWY
EAST
RAMP
DAVIDST
778
601070
BLACK
751060
GULLY
RD
5546
1657
282
BANGHOLME
RD
751080
POUND
MAXWELLCRT
RD
LONSDALE
100
1001080
10
80
CRESENT
WEDGERD
AMA
YLN
CRESENT
WOODBINERD
DORERD
CRANBOURNEROAD
INTERSECTION
NARRE
2WARREN
WARREN
RD
GRANDVIEW
TOORAKAVE
GROVE
DORSET
RD
RD
DORCAS
ST
DIANAST
BAYSWATER
TOLlMERIN
KIRKHAM
RD
AVE
TATTERSON
GRELAND
ST
RD
BERENDS
REDGUMDVE
NICHOLAS
DVE
DVE
QUINNST
WEBSTERST
FORSTERST
DANDENONG
GLADSTONE
DAYST
901080
RD
VALLEY
HWY
DAVIDST
PERRYRD
KIRKHAMRD
HEATHERTON
BRADYRD
BLACK
GULLY
RD
ARCHERRD
AVANDIA
CRESENT
ROAD
BLIND
INTERSECTION
BIGHT
1WEST
STATION
MICKLEST
MULGRAVE
FWY/SOUTH
EASTERN
ARTERIAL
HAMMONDRD
LANDEXCRT
BROAD
GULLY
RD
AQUEDUCTRD
CIVICDVE
GREENHILL
RD
BAXTER
TOORADIN
RDRD
(LANARCH
RD)
(metres)
SKYE
RD
(0.55KM
SOUTH
OF
...)
GRANTS
DION
AVE
(0.2KM
(0.35KM
EAST
OF.)
OF.)
(.25KM
NORTH
EAST
OF.)
GREENS
DANDENONG
RD
(BRAESIDE
VALLEY
DANDENONG
HAMMOND
RD
RD)
(0.4KM
WEST
OF
RD)
...)
CORNELlUS
ST
(0.07KM
WEST
OF
...)
(PRINCESS
HWY
EAST)
LONSDALE
ST
STUD
(DANDENONG
HWY)
RD
VALLEY
HEIDELBERG-KINGLAKE
RD
(DIAMOND
CREEK)
(DIAMOND
CREEK)
DABCAR
CLOSE
(0.7KM
EAST
OF.)
TOORADIN
CRANBOURNE
STATION
NARRE
NARRE
RD
WARREN
(O.4KM
SOUTH
RD
RD
(0.3KM
(0.25KM
WEST
OF.)
OF.)
OF.) RD)
FRANKSTON
RD
(CRANBOURNE
OF.)
RD)
QUARRY
NORTH
EAST
OF
...)EASTEAST
PRINCESS
HWY
EAST
(LONSDALE
ST)
MOUNTAIN
HWY
(WANTIRNASASSAFARAS
RD)
HWY
(FRANKSTON
DANDENONG
HENRY
RD
(0.31
KM
EAST
OF.)
CORN
Ell
US
ST
(0.07KM
EAST
OF.)
DANDENONG
VALLEY
HWY
(FRANKSTON-DANDENONG
COLEMANS
RD
(1.25KM
SOUTH
OF
...)
ZONE
SPEED
MELBOURNE:
SPEED ZONE CHANGEYANYEAN
SITES
USED
IN
EVALUA
TION
928
869
4234
2793
1156
1144
1884
484
GREENAWAY
ST
231
4761
2259
2028
2424
1590
1776
1970
608
4332
2177
2481
1422
1124
1748
2038
410
PARKRD
RD
MURNDALDVE
NEWMANS
60
RD
1354
1794
855
777
LENGTH
3259
599
748
MANNINGHAM
RD
BRIDGE
ST
516
912
225
581
ANDERSONS
CREEK
RD RD
KIMST
LANDSCAPE
DVE
2040
1733
HARRIS
KNEES
RD
GULLY
RD
HADSARRD
882
658
2505
1066
1203
238
713
985
678
SAINT
BAN
FOOTEST
BEAUTY
BRIDGERD
FOOTE
BULLEEN
NONS
HELENA
ST
GULLY
RD
RD
WARRANDYTE
CHURCH
BRD
RINGWOOD
MAXINEDVE
MINTER
ILMACRT
to
GULLY
70
60
90
CRT
RD
RD
RD
-RD
WARRANDYTE
RD
RD
3173
683
HIGH
DONCASTER
ST
BLACKBURN
RENOIRAVE
TINDALS
THOMPSONS
REYNOLDSRD
STINTONSRD
SPRINGVALE
RINGWOOD
RD
- RD
WARRANDYTE
TINDALS
BARADINE
EASTERNFWY
REYNOLDS
HELENE
RD
ST
RD
TCE
RD
WARRANDYTE
OLD
CHERRY
PORTER
WARRANDYTE
ST
TREE
RD
RD
W1LLlAMSONS
BLACKBURN
MANUKARD
NEWTONAVE
ENFIELDAVE
100
75
to
to
60
80
RD
1204
KURRAKRD
THE
YANYEANRD
RISE
80
1121
W1LLlAMSONS
RD
MITCHAM
CAROLE
80
RD
ROAD
RYANSRD
MITCHAM
CHANGE
RD
ROAD
NAME
ROAD
CALENDONIA
SPRINGVALE
ROAD
INTERSECTION
INTERSECTION
DVE
1OLD
2 WEST
720
DELPHIN
WALLOWAN
LAURIE
HUNTINGFIELD
75
ST
CRESENT
DVE
WATTLETREE
RD
MINTERCRT
RYANSRD
6324
BANNONS
HALEYS
FALCONER
FITSIMMONS
GULLY
LANE
RD
RD
LANE
RD
AND
ELSEMANS
CHARLES
BINGLEY
FAWKNER
CREIGHTON
BANNONS
DUGONGCRT
HURSTRD
HARRIS
KNEES
SUMMERHILL
Other
ERSON
RD
GULLY
AVE
STURT
CRESENT
LANE
ST
WAY
RD
DVE
GEORGEST
RD
DON
CASTER
RD
STATION
HARCRTST
RENOIRAVE
STITRAM
JOHANSONS
THOMPSONS
OLD
MANNINGHAM
TORTIADVE
WARRANDYTE
RD
RD
RD
ANDERSONS
CHERRY
BARADINERD
O'BRIENS
HARRIS
GULLY
TREE
LANE
CREEK
RDI
MANUKA
RD
RD
ARM
KANGAROO
STRONG
GROUND
RD
RD
ASHWOODDVE
FOOTEST
MELBOURNE
HILL
RD
AQUEDUCTRD
STCLEMS
ST
WARRANDYTE
PEMBROKE
BANKRD
STINTONS
RD
KENDALLS
MACKELROY
LANE
DVE
VICTORIAST
MOUNTAIN
HARCRT
VIEW
RDI
HALEYS
GULLY
RDI RD)
KENDALLS LANE
DONCASTER
RD
LANARCH
RD
TEMPLESTOWE
RD
(HEIDELBERG
-ST
WARRANDYTE
RD)RD
(MAIN
HURSTBRIDGE
RD)
HEIDELBERG-KINGLAKE
(DIAMOND
CREEK)
(RYANS
RD)
NEWMANS
RD
(HEILDELBERG-WARRANDYTE)
(metres)
W1LLlAMSONS
GREENAWAY
ST
(O.4KM
(DONCASTER
OF
ELTHAM
...) RD
WATTLETREE
RD
(MAIN
ST)
ZONE
SPEED
MELBOURNE:
SPEED ZONE CHANGEWARRANDYTE
SITES
USED
IN
EVALUA
TION
338
1531
4735
2787
864
362
1341
646
732
958
728
869
236
689
4255
2681
378
336
470
483
713
RANKIN
NESS
LANE
ST
RD
2661
2098
1290
1039
1202
931
484
666
525
764
970
975
967
205
140
367
810
673
770
MARRIOT
LANE
LENGTH
KANGAROO
RESEARCHRESEARCH-WARRANDYTERD
MELTONHWY
WARRANDYTE
JOSLYN
GLYNNS
KARDINIA
MARBERTCRT
LOOKER
DVE
RD
DVE
RD
832
1082
1122
1783
1703
592
583
332
298
647
630
756
177
1001080
751080
601070
75
to
60
VALLEY
LORIMER
RD
RD
THE
WEERONAWAY
EUMARELLA
LINK
ST
W1TLEYRD
HEDGELYRD
HOMESTEAD
CALDER
DVE
HWY
TAYLORSRD
GREEN
BERGINSRD
CALDER
GULLY
FWY
GROUND
RAMP
RD
KANGAROO
SUNSHINE
CALDER
RED
GREEN
LORIMER
MALCOLMCRT
PROCTORST
CHURCHILL
KERRISON
SAGEAVE
KERRS
HOWARDS
SHIRT
RD
GULLY
PARK
FWY
ST
AVE
RD
GULLY
GROUND
PARK
RAMP
DVE
RD
DVE
RD
RD
ROAD
ELTHAMDONCASTERYARRA
ROAD
GLEN
NAME
ROAD
RDIFITSIMMONS
RD
BUTTER
SUTTON
GRAHAMRD
INTERSECTION
MANS
ST
RD
TRACK
972
1255
CHANGE
GREENSBOROGH
PETRIKAVE
BONFIELD
CEMETRYRD
PARKLAND
100
751060
1080
ST
RD
1(NORTH
HWYI
HEIDELBERGKINGLAKE
WHITLESEASYDENHAMRD
SHARPSRD
TUNNECLlFFE
SHARPS
KEILOR
PARK
RD
KINGLAKE
DVE
DVE
KANGAROO
RDI
SACKVILLE
MOUNT
NEWRD
TAYLORS
CHARLlBURY
FISHER
STUD
WEBBER
DRISCOLLS
MELROSE
GRANDVIEW
ELTHAM
FLOODS
PECKS
CAPEL
LIBERTY
BELLBIRD
BLOOMS
HUMEDVE
BIRCHAVE
CAMPASPE
NEWRD
FITSIMMONS
FULLARTON
VALlAS
GROUND
RD
Other
Olher
GANGELHOFF
RD
ST
RD
ST
GROVE
PLEASANT
YARRA
-RD
AVE
PARADE
RD
YARRA
DVE
LANE
ST
CRESENT
RD
CRESENT
GROVE
GLEN
RD
GLEN
RD
RD
RDI
WELLER
RD
THE
GOSFIELD
GLENI
RD
HEIDLEBERG
KINGLAKE
CHERRY
TREE
RD
COOLAROO
JOANCRT
DONGOLARD
MANUKARD
AVE
TANJILCRT
SUNSHINE
KANGAROO
CARBINE
WAY
AVE
GROUND
RD
DON
CASTERELTHAM
RD/FITSIMMONS
JAYSONAVE
LANE
BERGINS
RD
LEE
DAWSON
BANKSIA
SLATER
ST
PARADE
GROVE
BAMBARARD
OLD
BUTTERMANS
ROAD
HURSTBRIDGE
INTERSECTION
RD
2(RESERVIORSUNSHINE
AVE
TARELLADVE
BRADWELLST
HEIDELBERG-KINGLAKE
RD
(GREENSBOROGH
HWY)
HEALESVILLEKINGLAKE
(KEILORMOUNT
RD
MEL
(MOUNT
TON)
SLIDE
RD
WARRANDYTE
KEILOR-LAVERTON
RD
(KINGS
SAN
ANGELO
RD)
OF
..) RD)
OLD
CALDER
HWY
(KEILORMEL
TON
RD)
MELTON
HWY
(KEILORMEL
TON
RD)RD
KEILORLAVERTON
RD
(GREEN
GULLY
RD) RD) RD RD)
(metres)
ELTHAMOLD
CALDER
YARRA
HWY
GLEN
RD
(MAIN)
MELTON)
ELTHAM
YARRA
GLEN
RD
(MAIN
HEIDELBERG
-LANE
KINGLAKE
RD
(MOUNT
SLIDE
ZONE
1001080
751060
100
601070
60
751080
to
1080
70
SPEED
MELBOURNE:
SPEED ZONE CHANGESTUD
SITESSTRD
USED
IN
EVALUA
TION
3184
939
3720
1198
953
190
981
1135
480
741
7067
1560
249
ROAD
NAME
1448
4232
9288
3370
2394
2780
1266
3586
5956
1091
RD
1835
717
COLCHESTER
RD
MULGRAVE
FWY/SOUTHEASTERN
ARTERIAL
1978
9353
L1VERPOOLRD
DVE
1007
2460
WELLINGTON
NAPOLEAN
DYNON
RD
RD
RD
KELLETS
1817
1116
765
MAHONEYS
RD
WELLINGTON
RD
2377
332
FAIRBAINRD
1010
246
496
964
788
1513
564
2096
2250
5491
FERNTREE
GULLY
3002
1359
LENGTH
ROAD
CHANGE
WANTIRNARD
3462
BORONIARD
DORSETRD
1578
FERNTREE
BURWOODHWY
HIGHSTRD
DORSETRD
HIGH
GULLY
RD
MORACKRD
NORTONS
RD
LANE
KELLETS
HWY
BORONIARD
SCORESBY
STUD
HWY
MOUNTAIN
MILLER
RD
WANTIRNARD
SELKIRKAVE
EYCOTST
FOOTSCRAY
FOOTSCRA
Y
SI
DUHLENBURG
SIMS
PLENTY
MS
RD
ST
KINGSBURY
HAVERBRACK
WALDEARD
751060
RD
DVE
HODDLE
DOCKLANDS
ST
HWY/CHARLES
ERIN
VICTORIAST
LORIMER
GRIMES
ST
KEONPARADE
NAPOLEAN
RD
RD
SOUTHEASTERN
FWY
FELSTEAD
AVE
MAHONEYSRD
SHEAR
GOLD
CRT
FAIRBAIRN
RD
BOUNDARYRD
FORRESTST
WESTST
FOLEYSRD
933
WELWYN
601080
PARADE
McOONALDRD
SAINT
ALBANS
RD
868
469
SOMERVILLE
TILBURN
SUNSHINE
RD
AVE
RD
1491
639
931
946
3935
751060
JACKSONS
ROAD
ROAD
INTERSECTION
INTERSECTION
1DVE
2
OLIVE
DORCASST
GROVE
DVE
SCORESBYRD
MILLERRD
NAPOLEANRD
MOUNTAIN
WATERDALE
BURWOODHWY
PIGOTTST
HOBSONS
RADCLlFFE
EPSOMRD
RD
ST
PAVIT
LANE
LYSTERFIELD
JOHNSON
DUNSTANS
ST
CRT
RD
SOMERVILLE
GRANGE
HODDLE
ST
RD
BLAXLAND
TILBURN
RD
NEALEAVE
INDUSTRY
PARK
RD
GILMOURRD
FURLONGRD
WESTERN
McDONALD
MARKETRD
Other
RING
RD
RAMP
FITZGERALD
ADELAIDEST
ST.
ALBANS
RD
ANDERSONRD
601070
BIRDST
LEILAST
GLENCAIRN
BIRD
ST
AVE
ADELAIDE
751080
ST
ROBINSONS
RD
BLACKBURN
RD
WAVERLEYRD
NORMANBYRD
HUNTINGDALE
RD
ACACIARD
STUDRD
HOLMES
ST
FOLEYS
RD
BALMAIN
ST
COOTAMUNDRA
ELM
STEAD
DVE
FWY
RAMP
FARNSWORTH
ST
(STATION
RD)
(metres)
PRINCESS
WEST
RD)
PIGOTT
EPPING
(NORTH)
ST
RD
(SOUTH
(HIGH
ST)
OF
...)
KEILORLAVERTON
RD(SMITHFIELD
(BALLARAT
(MOUNT
DERRIMUT
RD)
MAIN
RD
(MAIN
RD
EAST)
KEILOR-LAVERTON
RD
(STATION
WESTERN
HWY
RD)
WESTERN
HWY
(BALLARAT
RD)
MONASH
HWY
(WELLINGTON
RD) RD)
THATCHER
RD
(0.5KM
WEST
...)
ZONE
SPEED
MELBOURNE:
SPEED ZONE CHANGEWARRIGALRD
SITES CHANGE
USED
INWELLINGTON
EVALUATION
792
4961
1523
1933
318
346
531
484
236
264
328
56
1024
515
359
FWY
RAMP
FERNTREE
AURISHAVE
HIGHBURY
RD
GULLY
RD
HIGH
STRD
75 to
60
to
80
60
LENGTH
ROAD
MULGRAVE
SPRINGVALE
ROAD
RD
RAMP/SOUTHEASTERN
NAME
WAVERLEYRD
FAIGH
ST HWY
ARTERIAL
MONASH
HWY/
GARDENRD
HAVERBRACK
RD
DVE
SPRINGVALE
JACKSONS
MULGRAVE
FWY/SOUTHEASTERN
RD
ARTERIAL
BLACKBURN
RD
ROADWELLINGTON
ROAD
INTERSECTION
INTERSECTION
1 RD (DANDENONG
2
70
60
(metres)
PRINCESS
EAST
RD)
ZONE
60 to 70
SPEED
APPENDIXD
Listing of speed zone change sites evaluated in the rest of Victoria, length of
road covered by each speed zone change and estimated crash effects by speed
zone change type within LGAs
nge*
RURAL VICTORIA
Lower 10644.42%
374.98%
Limit
117.39%
1408.64%
-95.40%
-96.77%
-98.02%
0.9018
-16.66%
-44.83%
-60.00%
-96.57%
0.3683
-40.00%
-96.71%
0.1221
0.4419
0.9153
-63.89%
-92.57%
-96.85%
-95.91%
-16.66%
99.99%
-17.86%
80
-54.54%
-75.73%
0.3045
60
384.62%
-2.78%
44.44%
to
Limit
2927.69%
-95.41%
-79.99%
0.4817
146.16%
28.57%
2460.95%
254.72%
-64.65%
-77.52%
-81.55%
0.8721
0.5372
140.01%
-10.71%
41.03%
3564.07%
1357.56%
950.63%
842.70%
707.18%
566.48%
-92.29%
-96.14%
-97.55%
-88.07%
-99.40%
0.6813
0.5500
0.7265
0.8493
0.8118
0.9004
0.9227
0.6136
0.7120
-30.00%
-25.00%
-80.00%
109.09%
14.28%
2314.79%
1498.79%
1901.72%
756.08%
296.19%
-93.75%
-95.49%
86.24%
36.25%
-93.20%
-91.55%
-92.73%
-96.28%
-72.86%
-99.27%
0.4876
0.4046
0.8866
1.0000
0.8047
0.5474
200.00%
0.00%
-28.90%
-75.00%
-68.52%
-20.00%
0.00%
42.86%
16.67%
2443.58%
7724.68%
8370.48%
2354.92%
1286.58%
1570.01%
1141.13%
1863.01%
1599.58%
1190.24%
-93.79%
109.29%
-95.63%
-94.89%
-92.66%
-95.39%
-95.76%
-95.13%
-98.67%
-77.47%
-85.54%
0.1652
0.9897
0.7579
0.8915
0.8983
0.9057
0.8567
0.9287
0.8070
0.6611
0.4409
0.8306
75
37.51%
250.01%
-83.33%
-22.50%
-37.50%
-29.55%
1.85%
9196.32%
2523.07%
5066.10%
5662.08%
436.52%
-82.38%
-92.05%
-87.66%
0.9742
0.1912
0.5316
0.6414
0.8037
166.66%
1856.35%
-98.72%
0.7110
3100.01%
2078.34%
-89.50%
0.6778
0.8618
100
83.33%
to 70
80
4539.08%
462.69%
-87.58%
0.6293
0.5623
2807.35%
1594.66%
1751.56%
1217.86%
1985.57%
-93.57%
-89.99%
0.6057
0.9520
-53.85%
20.00%
70.59%
9.09%
2117.48%
1948.69%
1619.63%
-92.55%
-91.62%
0.8828
166.66%
23.53%
1069.85%
1008.79%
1770.95%
166.58%
-93.10%
-97.75%
0.6859
-50.00%
-22.86%
13.63%
3595.05%
1189.92%
-89.18%
-98.06%
0.6760
-50.00%
769.44%
Upper
Pr>
Ch-Sq
75 to 80
ALBERTON
Percentage
RURAL VICTORIA
Percentage
Change*
Lower
Limit
Upper
Limit
Pr> Ch-Sq
242.84%
-30.00%
-50.00%
33.34%
183:35%
-50.00%
-25.00%
150.00%
-66.41%
-94.45%
-96.26%
-91.41%
-84.78%
-98.72%
-96.79%
-90.02%
3399.10%
782.15%
568.45%
1970.81%
5174.28%
1856.35%
1650.71%
6161.29%
0.2985
0.7826
0.6003
0.8371
0.4851
0.7110
0.8579
0.5771
Speed Zone Change Within Each LGA
WAN NON
WONTHAGGI
75 to 80
75 to 100
Other
75 to 80
75t07O
100 to 80
75Tci8O
60 to 80
*: Net percentage change after adjusting for control group crash trend
RURAL VICTORIA:
LGA NAME
70
100
to BY
80 SPEED
2424
1077
3594
75
to
80
748
554
753
983
860
1002
726
3553
2073
5790
3164
2117
1029
3272
684
694
278
924
1188
506
561
1556
1941
4922
304
4786
60
1988
5416
1786
1327
956
11061
4632
3443
3069
668
1680
380
886
781
388
636
SPEED ZONE
CHANGE
LENGTH
OF ROAD
CHANGE
TOTAL
ROAD
LENGTH ZONE
(metres)
IN EACH LGA
0
RURAL VICTORIA: LENGTH
OF
ROAD
ZONE
CHANGE IN EACH LGA
10290
4720
1045
1960
3764
349
70
556
3314
4251
4256
932
5181
2893
998
560
379
2110
2810
75
475
421
771
861
546
to
80
1575
3465
1780
255
342
1274
516
1605
481
60
10569
1804
1170
SPEED ZONE
3204
CHANGE
5618
991
100
3212
8881
toBY
80SPEED
TOTAL
ROAD
LENGTH
(metres)
60 to 70
LGA NAME
RURAL VICTORIA: TOTAL ROADTOTAL
LENGTH
BY LENGTH
SPEED ZONE
CHANGE
21699
99171
26647
9242
5199
7096
10041
12249
ROAD
(metres)
100
75
75
Other
to
100
80
60
60
to 60
80
70
75 to 70
SPEED ZONE CHANGE
RURAL VICTORIA:
ZONE
CHANGE
SITES
IN EVALUATION
2959
417
1988
594
1275
562
651
402
267
748
700
722
943
1929
781 SPEED
1240
534
304
760
824
202
830
658
733
DICKENS
RD
312
1217
536
186
560
351
391
447
432
952
531
860
563
561
498
394
622
647
293
715
825
625
210
683
516
1117
1187
761
998
627
929
WANDANA
BASSHWY
DVE
970
HENDY
MAIN
CAPE
MERRIMANS
HYLAND
OTWAY
HWY
CREEK
RD
RD
CHANGE
BETLEY
ST
LENGTH
ROAD
ROAD
NAME
280
BARRABOOL
554
500
1354
AGAR
RD
NORSEMANS
RD
AGARRD
YARRINGARD
BENDIGO
--HIGHWAY
RD
BAMBRARD
ST
ARNAUD
ST
75
to
100
GREAT
SOUTH
OCEAN
GIPPSLAND
RD
HWY
POUND
RD
EAST
GRANTVILLE
GLEN-ALVIE
RD
BASSHWY
2OF
THE
STRAND
HYLAND
ST
ELSO
75
RD
HWY
CRESENT
--(KEERA
INTERSECTION
BRIDGEWATER
GHAZEEPORE
to
80
RD
DONOLL
MOLlAGUL
EDINGTON
ROAD
ROAD
DUNOLL
Y
RD
ST
ARNAUD
DUNOLL
YROAD
RD
FERRIER
ROAD
HONOUR
WILLOWBANK
ROAD
HAMILTON
HIGHWAY
GLENELG
CUTTY
SARK
RD
BARRABOOL
RD
JAMES
COOK
DVE
OLD
CAPE
CUTTY
COACH
OTWAY
SARK
RD
RD
PATON
RESERVIOR
ST
RD
GLADSTONE
YARRINGA
RD
NEWBRIDGE
-ROAD
TARAGULLA
RD
BOUNDARY
ROAD
AIREYS
ST
HIHOS
ELGIN
LANE
SHUNTOFF
BANKS
MERRIMANS
HENDY
RD
EXTENSION
CREEK
MARTINS
RD
RD
ROAD
INTERSECTION
1
ANGELSEA
LYTTON
ST
WAYMAN
ST
DONOLL
Y
ROAD
RAGLAN
STREET
CARDIGAN
STREET
MARYBOROUGH
-OF
DUNOLL
ROAD
CLARK
STREET
ORMEST
COCHRANE
ST
CANDOVER
70
ST
BELL
PDE
MONT
ALBERT
RD
60
KILMOUR
ROAD
THE
BOULEVARD
75
HOWEY
STREET
McBEAN
AVENUE
Other
GISBORNE
McGEORGE
MEL
TON
ROAD
MOUNT
GISBORNE
ROAD
HIGHWAY
CASTERTON
-RD
PENOLA
ROAD
100
to
80
BETLEIGH
ST
LAANECORRIE
•AITKEN
NEWBRIDGE
RD
HOSPITAL
STREET
WELSH
BETLEY/
DUNNOL
Y
-ROAD
EDDINGTON
ROAD
MARYBOROUGH
-DUNOLLY
DONOLL
YUSED
ROAD
MONT
ALBERT
RD
BACCUS
MARSH
ROAD/
ROBERTSON
STATION
HAMILTON
ROAD
ROAD
STREET
MOUNT
MACE
DON
ROAD
WEST
OF
MONAHAN
GISBOURNE
ROAD
MONAHAN
ROAD
-ST
KILMORE
ROAD
FRITH
ROAD
STREET
PANORAMA
FREEWAY
CALDER
MACEDON
FREEWAY
RAMP
GAVIN
ROAD
SERVICE
NORTON
ROAD
HOWEY
STREET
CHAFFREYS
LANE
GLENELG
60
HIGHWAY
MOUNT
ELEPHANT
LLOYDS
to
ROAD
80
MOUNT
GAMBlER
ROAD
(metres)
WATERFALLS
CALDER
ROAD
BACCUS
MARSH
-ROAD
GISBORNE
ROAD
/EAST
ROBERTSON
STREET
RODNEY
STREET
WILLOWBANK
NORTH
OF
COUANGAL
ROAD
T
GISBORNE
BRADY
ROAD
-RD
MEL
TON
ROAD
GISBOURNE
ROAD
CASTERTON
PENOLA
ROAD
(SOUTH
...
)-)WEST
ST
ARNAUD
--ROAD
DUNOLL
Y
RD
(MAIN
ST)
ELGIN
ST
(WEST
...
CARPENTER
ST
LYTTON
(ACROSS
ST
INTERSECTION)
...
OF
)OF
BRIDGEWATER
DUNOLL
YROAD
RD
(COMMERCIAL
RD)
RAGLAN
STREET
(NORTH
WEST
OF
...
)) (SOUTH
HWY
(MELBOURNE
RD)
MONOHAN
ROAD
(EAST
OF
...OF
)...
GRANTVILLE
-RDERVINS
GLENALVIE
RD
(SW
&
NE
...
)...
DUNNOLL
Y-EDDINGTON
RD
(SOUTH
OF
)OF
DUNOLLY
ROAD
(ELGIN
STREET)
RAGLAN
STREET
OF
OLD
BEALlBA-MOLlAGUL
RD
(SOUTH
OF
...
)Y
AVOCA
-MAIN
BEALlBA
CLARKE
RD
(NORTH
STREET
WEST
(SOUTH
OF
...
)EAST
) OF ... )
PRINCESS
ST/
LA
TROBE
TERRACE)
GISBORNE
ROAD)
KILMORE
(SAUNDERS
--•MOLlAGUL
KILMOUR
(STATION
ROAD)
BARTAGUNYAH
(EAST
OF
...
)(NORTH
(HAMILTON
STREET)
MELBOURNE
(HOWEY
STREET)
HOWEY
STREET/
GISBOURNE
MELBOURNE
COUANGALT
(SOUTH
...
) )...
WILLOWBANK
ROAD
(SOUTH
OF
...
BARTAGUNYAH
ROAD
(NORTH
EAST
ROAD
ZONE
ROAD (MELBOURNE
ROAD)
100
to100
80
60
75
to
75Other
to
80
SPEED
RURAL VICTORIA:
ZONE
CHANGE
SITES
USED
IN EVALUATION
752
924
1345
441
2082
424
651
979
1047
1189
640
1031
1222
470
679
481
267
497
448
856
1385
666
1384
506
338
LENGTH
779
694
388
290
542
545 SPEED
728
1804
2117
313
1135
2328
561
1074
910
371
394
886
726
1083
1300
263
550
1328
361
288
346
370
789
333
75
to
80
595
ROAD
500
HAMILTON
HIGHWAY
278
CHANGE
807
HENTYHWY
380
355
SERVICE
MORISON
ST
CEMETARY
CEMETERY
RD
ADAMS
1NTON
RD
ROAD
GREAT
NEYLON
OCEAN
ST
RD
PRINCETOWN
RD
STRZELECKI
ROAD
NAME
LAMBERT
KELLETT
FLACKS
HOLLlS
RD
60
70
COLAC
PRINCESS
-RD
BALLARAT
HWY
TINAMBA
CO
LAC
SEATON
BELL
ST
100
to
L1COLA
COMMERCIAL
MCMILLAN
FIREBRACE
--SALE
MAFFRA
STREET
RD
RD
RD
TEAS
MAFFRA
HEYFIELD
BOISDALE
STRATFORD
-STONEYFORD
ST
INVERLEIGH
-HWY
RD
RD
RD
Other
TEASDALE
FULTON
RD
LETHBRIDGE
LAVERSHILLS
CAMPER
DOWN
PORT
-MAFFRA
COBDEN
-ROAD
COBDEN
RD
RD
RD
WALKERST
TOMAHAWK
GREAT
COBDEN
OCEAN
TERANG
CREEK
RD
PRINCETOWN
RD
ST
RD
COB
DEN
RD
CAMPERDOWN
DESAILY
-CLARKE
STW
COBDEN
ROAD
COBDEN
TERANG
RD
LAVERS
COB
DEN
HILL
-RD
PORT
-ROAD
COBDEN
CAMPBELL
RD
RD
NORTH
OF
ROBIN
TIMBOON
VAGGST
SONS
-80
BRUCKNELL
RD
BALDING
ST
BATHS
RD
TIMBOON
NULLAWARRE
PORT
CAMPBELL
RD
RD
TIMBOON
ROAD
INTERSECTION
-ST
BRUCKNELL
GLERUMS
ROAD
INTERSECTION
RD
RD
PORTLAND
-STATION
NELSON
BLACK
PRINCESS
SWAN
HWY
RD
OAKBANK
RD
HAIGH
ST
EXTENSION
OLD
SALE
CAVE
HILL
RD
BOYERS
RD
MOE
-DALE
GLEN
GARRY
RD
70
CARRS
RD
LANE
ORCHARDS
RD
WOOLSTHORPE
-CAMPBELL
HEYWOOD
RD
HALLS
FERRERS
HILL
RD
RD
CAMERONS
MELLlNGTONS
LANE
RD
TINAMBA
ROSE
MAHERS
ST
LANE
-RD)
SEATON
RYANS
DREWST
DEANS
RD
RD
TRARALGON
--RD
BRIAGOLONG
-HWY
MAFFRA
RD
ROAD
RD
MERRY
McADAM
WEIR
RD
ST
ST
BREWERS
FULTON
75
to
100
HILL
SHELFORD
BANNOCKBURN
MALMOST
CAMPBELL
BAKERS
LANE
STRATFORD
HEYFIELD
RD
-OF
SEATON
-RD
MAFFRA
RD
RD
80
CASTLE
OLD
THORPDALE
ST
RD
RIX
AVENUE
ELIZABETH
CRT
GIBSON
ST
ELDON
CRT
COUPER
BOOLARRA
AVE
/+RD
INVERLEIGH-SHELFORD
MURRAY
McKINNONS
RIVERSDALE
ST
RD
TRARALGON
RIVER
STREET
--...
MAFFRA
RD
ASH
ELEANOR
ST
COMMERCIAL
THOMPSON
MERRYST
ST
RD
MORONEYST
SCARLETT
ST
SHELFORD
-ROAD
BANNOCKBURN
RD
FARMER
WEMBRIDGES
RD
/ RD
WANKE
OLD
SALE
RD/
TORRES
ST
GALVINS
FLORENCE
AVENUE
BOOLARRA
SOUTH
MIRBOO
NORTH
ROAD
(metres)
WALKER
ST
(EAST
OF
...
)2
CURRELLS
RD
DIGNEYS
BRIDGE
COBDEN
(SOUTH
-(.8KM
PORT
CAMPBELL
DEN
OF)
(SOUTH
(WEST
OF
) ) ... )
CAMPERDOWN
LAVERS
-L
DERRINALLUM
HILL
COBDEN
ROAD
RD
OF
...
...
)) ...OF
ROBINSONS
(NORTH
OF
...
)EAST
WADE
STREET
(SOUTH
EAST
OF
...)
ROKEWOOD
SKIPTON
RD
(FERRARS
WEIR
RD
OF
...
)WEST
WESTBRIDGES
/RD
WANKE
RD
(NORTH
STRZELECKI
(MIRBOO
NORTH
-ALLlTS
THORPDALE
RD)
HENTY
HWY
(WEST
OF
...
)1
WILSON
STREET
ELEANOR
(SOUTH
OF
STREET
...
)COB
(.56KM
NORTH
WEST
OF
...
HOPES
PLAIN
RD
MERRY
FIREBRACE
(NORTH
STREET
RD
)EAST
(NORTH
(1.16KM
WEST
NORTH
OF
OF
...
)OF
) )ST)
BAKERS
LANE
(1.39KM
EAST
OF
...
)...
MEENIYAN
--COLAC
MIRBOO
NORTH
RD
(OLD
MARDAN
RD)
HWY
(RIDGEWAY)
JOHN
FIELD
DVE
(MOE
-GLEN
GARRY
RD)
PRINCESS
FREEWAY
(WEST
OF
...
)SOUTH
-RD
ROAD
(BAROMI
MURRAY
(.4KM
EAST
OF
...
)EAST
MERRYDALE
STREET
(SOUTH
OF
...
) ...
SCARLETT
ST
(SOUTH
OF
...
)RD
STRZELECKI
SRTZELECKI
(SOUTH
WEST
HWY
OF
(SOUTH
)(1.5KM
OF
...
ZONE
75
60
100
to100
80
75
75to
to
80
70
SPEED
RURAL VICTORIA:
ZONE
CHANGE
SITES
USED
IN EVALUATION
781
941
180
349
379
1154
471
1522
498
453
1804
921
2726
1170
1036
786
206
2029
736
561
501
636
979
546
LENGTH
1405
1305
447
1245
510
560
1450
821
230
525
1504
569
716
266
474
850
766
895
242
1756
982
283
556
550
689
421
2393
2252
771
671 SPEED
1802
600
715
606
MACKEYS
CRINIGAN
WATERLOO
WATSONS
HAMILTON
HIGHWAY
RD
SWITCHBACK
1693
ROAD
MOE
--YANG
GLEN
GARRY
RD
255
GRAMPIANS
ROAD
-•RD
-ISLAND
FORREST
COLAC
RD
228
GREAT
RD
SETTLEMENT
BOOLARRA
L1MONITE
-HIGHWAY
BOOLARRA
-TRARALGON
CHURCHILL
MARINE
LYNNE
STREATHAM
MACKEYS
AVENUE
RD
RD
MONASHWAY
GRANYA
GVE
MORWELL
--RD
MARYVALE
RD
RD
MOFFAT
ST
BARDIA
75to
OLD
RD
WATSONS
NORTHERN
AVE
ROAD
NAME
VENICE
YORKST
GLENELG
WIGANS
LANE
BIRREGURRA
FORREST
RD
BEECH
FOREST
HENRYST
GELLlBRAND
GELlIBRAND
-RD
CARLISLE
RD
MOOMOOWROONG
RD
75
COGHLAN
COWES
-ISLAND
RHYLL
RD
GROSSARD
RD
LYALL
RD
PHILLlP
RD
PHILlIP
RHYLL
-OCEAN
NEWHAVEN
RDRD
VENTNOR
BEACH
FAIRMONTST
WATTLE
CRES
CANTEBURY
MATTA
HEXHAM
-ST
CHATSWOTH
CRINIGAN
SYMONSCRES
PRINCESS
FWY
MORWELL
CENTRE
FARRAN
SWAN
RD
-)RD
THORPDALE
RD
WOOLSTHORPE
GRAEMEST
STREATHAM
RD
VINCENT
GRAEME
RD
MANNING
DVE
LANGFORD
AUSTIN
BENNETT
ROAD
INTERSECTION
ST
INTERSECTION
1
2
ALBERT
ST
MAHONEYS
PENSHURST
-(SOUTH
DUNKHELD
ROAD
McCLEOD
STREET
POWLlNG
BALBEGGIE
LANE
DICKLE
STREET
BOUNDARY
ROAD
DEPOT
ROAD
HENRY
ST
PENHURST-WARRNAMBOOLROAD
60
to
70
100
to
80
COWES·RHYLL
RD
HARBISON
RD
Other
80
60
to
75
to
100
WYNNE
RD
FRANKLYN
ST
COGHLAN
LYALL
McDONALD
WAY
WINCHESTER
WAY
ARMITAGE
FLORENCE
STREET
AVE
DONLAD
FORBES
STREET
GRAMPIANS
ROADI
VICTORIA
VALLEY
ROAD
SOUTH
OF
RITCHIE
STREET
FRIZON
COLAC,
COLAC
-AVE
LAVERS
HILL
RD
SETTLEMENT
RD
SILVER
LEAVES
DUNSMORE
RD
PHILLlP
ISLAND
RD
GRIFFITHS
ST
GRAYDENS
RD
FISHER
ST
RHYLL
-DVE
NEWHAVEN
ROSE
RD
ST
BERMAGUI
CRES
VENTNOR
BEACH
RD
VENTNOR
COWES
-RD
BEACH
COWES·
RD
RHYLL
RD
IBAY
VENTNOR
BOYS
HOME
RD
HASTINGS
MCHAFFIE
ST
(metres)
GElIBRAND
RIVER
RD
LOY
-SALE
-POINT
MORWELL
RD
(COMMERCIAL
RD)
COMMERCIAL
RD
(MORWELL
-CHANGE
THORPDALE
MOE
GLEN
GARRY
RD
(NARRACAN
DVE)
MOORE
ST
(MOE
- RD
RAWSON
ST)
FORREST·
APOLLO
BAY
RD
(GRANTS
ST)
PHILLlP
RD
(BRIDGE)
OF
CHURCHILL
TRATALGON
RD
(TRAM
WAY
RD)
MARYV
ALE
RD
(ALEXANDER
RD
I TRAMWAY
PRINCESS
RD)
FWY
LOY
RAMP
-ST)
YANG
(EAST
-(WEST
MORWELL
...
LANE)
MORWELL
RD
(pRINCESS
HWY)
GRAHAM
CRINIGAN
STREET
RD
(NORTH
OF
OF
)(FIRMINS
...
POWlING
LANE
POWLlNG
(WEST
LANE
...
)WAY
(EAST
OF
TI
TREE
LANE
BATESWORTH
OF
...
)ST
LANE
(SOUTH
OF
... )(0.5KM
)OF
LARDINER
RD
(SOUTH
BEECH
-OF
FORREST
RD
(5KM
NORTH
OF
SKENES
RD
(EAST
OF
...))RD
)RD
FAIRMONT
ST
(SOUTH
WOOLSTHORPE
--IROAD
STREATHAM
RD
(WEST
OF
...
RD
BIRREGURRA
-(NORTH
FORREST
(WEST
OF
...
)...
RD
OF
GAMBlER
...
FORRESTAPOLLO
FORREST
BAY
RD
-LANE
APOLLO
OF
...
)RD
RD
(EAST
......) )WEST OF ... )
GRIFFITHS
RD
(SOUTH
EAST
OF
...
SKENES
CREEK
RD
(WEST
OF
...
)AVE
SETTLEMENT
RD
(ACROSS
INTERSECTION)
MCILLWRAITH
(REID
ICREEK
RHYLL
- )(WEST
NEWHAVEN
ZONE
75
80
60
to
70
100
to100
80
Other
75
to
SPEED
RURAL VICTORIA:
ZONE
CHANGE
SITES
USED
IN EVALUATION
716
440
475
2227
514
769
771
75
to
100
557
694
1114
2269
871
423
403
2424
870
LENGTH
ROAD
511
342
BASSHWY
CHANGE
1703
255
407
ALBERT
RD
TOOLONG
RD
1045
861
416
1912
710 SPEED
MURRAY
RD
MURRAY
FELLOWS
RD
WARD
MCKENZIE
VENTNOR
751070
751060
RD
RD
RD
319
751080
WADECRT
1790
FULLER
RD
1001080
75
to
80
MAFFRA
-ROAD!
RD
PRINCESS
HWY
GLEN
SOMERTON
CAMPBELL
PARK
CRT
RD
834
HWY
GLEN
CAMPBELL
OTWAY
GRASS
CRT
7510100
DALE
ST
RD
516
COBAINS
RD
601080
RD
1196
573
317
SOUTH
GIPPSLAND
HWY
WANDANA
REEVE
ST
RD
JOHNS
60
ST
to
70
400
180
KANGAROO
ROAD
REGENT
STREET
780
Olher
3032
1096
2213
613
278
MAN
DU
RANG
GULLY
ONR
SINCLAIRS
100
TREE
to
80
HILL
ROAD
1082
1509
190
NANKERUS
ROAD
TROTTING
TERACE
TANNERY
977
2685
705
BASS
HWY
TEMPLETON
STREET
CARNEYS
RD
ROAD
NAME
ROAD
INTERSECTION
ROAD
INTERSECTION
1
2EAST
VENTNOR
RD
BEACH
RD
JUSTICE
NELSON
LYALL
IRVING
RD
RD
RD
KING
FELLOWS
ST
CAMERON
ST
RD
REED
CRES
PINE
GVE
481
CAPE
PATTERSON
DUDLEY
RD
GRAHAM
ST
BASS
GARDEN
HWY
ST
\ROAD
RD
DAWSON
ST
GIBSONS
HUNT
PLACE
PARK
ST
BENDIGO
BENDIGORAGLAN
SOMERTON
ST
-SALE
REDESDALE
REDESDALE
PARK
DVE
RD
STAWELL
PUNT
LANE
ST
DAWSON
ST!
MAFFRA
SALE
STORYS
ROAD!
MANNES
EMU
CREEK
ROAD
ROAD
SHEL
TONS
ROAD
601070
JUNORTOUN
MANDURANG
-REDESDALE
ROAD
ROAD
ROAD
TANNERY
SUGARLOAF
FRANCES
NORTH
OF
STREET
LANE
CAVAGNAS
COUSINS
ROAD
STREET
MclVOR
HIGHWAY
MANDURANG
ROAD
HAWKINS
LANE
TANNERY
LANE
RED
TANK
ROAD
BENDIGO
-REDESDALE
ROAD
MclVOR
HIGHWAY
M
clVOR
HIGHWAY
GORDON
STREET
H
IRDS
LANE
GLENELG
HIGHWAY
&
LOCH-WONTHAGGI
WATTST
STEWATRT
ST
KORUMBURRA-WONTHAGGI
RD
IVOR
ST
VICARS
AVE
COUSINS
ST
STORYS
EMU
CREEK
ROAD
BAKERS
LANE
TROTTING
TERACE
AT
COLERAINE
INTERSECTION
- ROAD
EDEN
BIW
HOPE
STH
ROAD
DUDLEY
AERODROME
RD
(RAGLAN
(INVERLOCH
RD)ST)
PRINCESS
HIGHWAY
(ROSEBANK
RD)
THISTLE
PLACE
(WEST
OF
...
)WHITE
BELLARINE
HWY
(QUEENSCLlFF
RD)
VICARS
AVE
(NORTH
OF
...
PARK
ST
(WEST
ST
(0.8KM
OF
...
SOUTH
)(EAST
EAST
PRINCESS
HWY
(YORK
ST)
STORYS
ROAD
(EAST
OF
...
)ST)
CAVAGNAS
SUGARLOAF
ROAD
ROAD
(SOUTH
(5KM
WEST
OF
...
...
TROTTING
TERRACE
OF
...
)...OF
CARNEYS
RD
(EAST
OF
...
) )OF
(metres)
SEAWARD
DVE
(0.32KM
NORTH
OF
...
)RD
AERODROME
RD
(RAGLAN
STOREY
ST
(NORTH
EAST
...
))OF
KAIRNS
ROAD
(0.38KM
KAIRNS
NORTH
(0.15KM
WEST
SOUTH
...)
OF
) )... )
STORYS
ROAD!
EMU
CREEK
ROAD
(WEST
...)
MURPHY
LANE
(WEST
OF
...)
ZONE
RD
60
to 70
751070
75
80
1001080
751080
SPEED
APPENDIXE
Chart of Definitions for Classifying Accidents (DCAs)
DEFINITIONS FOR CLASSIFYING ACCIDENTS
PEDESTRIAN
DN FDOT
IN TOY (PRAM
VEHICLES
FRDM
ADJACENT
DIRECTIDNS
(INTERSECTIONS
DNL Y)
--
VEHICLES FROM
SAME DIRECTION
IOPPDSING
VEHICLES FROM
DIRECTIONS
VEHiCLES
'~.::.:
'--"
..
I00
NE AR SIDE
CROSS
102
SID!
WQAAIIfG
'):
l YIIfG
OH CARRIAGEWAY
103
RIGHT
----- .. ::.:::
~:;.~:::
........
111
fAR
lEflfAA
,-w
ptAYING
"'
WllH
TRAffiC
TWO
...
.
.
..........
.
105
fACIHGfRAfFlC
.. cccccc.~
RIGHT
fiXED OMCT'
,
RIGHT
lEfT
115
fAR
RIGHT
RIGHT
lEflWT
REAR
I 0,
132
SIDE
SWlP!
,-,,---"
133
lANE
CHAIfGE
125
lANE
CHANGE
lEfT
135
RIGHT
OH fOOTPATH.
~;.~V··~
.....
.
·H"',·
.••.••••
H:·::::"'L:;":
H'
..................
"'
107
DRMWAY
-'----
S TRUC,
0r
J
.
lffTRIGHT
c;:;:J
WHIlE
fAR
BOARDING
OR ALIGHTIHG
VEHiCLE
108
TWO
lEfI
l.
TURH
';INITlfl"tC
rf""\n
r/""'lJ""'l"'''''''
127
SIDE
TURN
SWlP!
IN
144
.REAR
OR
COWN
1351
153 ..
VEHK:LE
DOOR
, SID!
WlTURN
SWlP!
137
ROAOWQR~S
BEND
'PARKED
INTO
172
1631
',I'"
154
•••m;.:..:..;..
........
.....
Of
fROM
VEHICLE
~~~
lOAD
13181
OR MISSilE
STRUC~
VEHICLE
BEND
'"
'~2
STRUCX
156
x
~1(8J
'"
STRl.(x
OB%~TlEtiR~\~O~",;~ll
C 183
CROSSING
RAilWAY
fURNITURE
SAME
~~~~~~ ,Fl~~~
14 7
~~
.
OUT
Of
PARKED
RUN
-.-
CAR
AWAY
CONTROl
184
194
175
185
195
176
185
196
177
187
197
174
ON CARRIAGEWAY
.'~
...~~::::::j
... .•
Off
155
EHO
Of
ROAD
T INlERSECilON
~~~~,~~~iy
156
1571
I~~~~;n)
157
OTHER
158
158
148
OTHER
OVERTA~ING
MANOEUVRING
DIRECTION
193
~cCf
:
OTHER
OTHER
OTHER
CROSSING
192
TRAiN
't
~j)
o~m~~~~no~~I~~E
146
FOOTWAY
191
vO
u'u
fROM
138
190
..
COHTROl
ON CARRIAGEWAY
IN
~crp1\
~~
17 31
R:~~6g~eW~~,~~t
0
•
0.1
CARRIAGEWAY
lEfT
18
'"
VEHICLE
:i~i·;'~::';::
'"
OUT
P!R~NWR~~~:~~~T1QN
RIGHT
OMCT
Off
TO RIGHT
>;~:;~/~
154\
J
~:
::<.
C
155
145
•••
t
CARRIAGEWAY
l..--J]
REVERSING
.................
Off
162
OUT
END
:-:.:.I ..:-:.: .. #~
..C:'.~:".::"
~:::;:Jl
1
---.....
PAR~INGVEHIClESOHlY
I~l~
128
118
119
•••••••••••••••••.•..••
-~
CUTIING
Off
171
FEll
t
CARRIAGEWAY
RIGHT B_EN_0 __
~::;~~E~~i~~:O
151IlE6iJ~~fl
PARKED
Off
~:
::;;~~;::;
r-
BRC~EN
170
TO lEfT
If
cO=
/;~
UL1JiJI/
.....................
ADJACENT
109
152
~
I_~
OTHER
OTHER
P\OlSTR'AH
125
117
OuT
I
y
---:\
L
115
~PlJllING
•
er
-~
DOUBLE
ACCiDENT
.L.-~
I
RfAR
~--C , .
PUlLING
1341
15 I
o
CARRIAGEW'Y
160
P1R~EO
CZJf2
l..--lZl
•
I
0<=>0
I~
lEfT
143
PAR~ING
CONTROl
Off
150
~:
.............. ..
T£MPORARY
~
105
ENT£RING
RIGHT
InolOvtrt.k""lJ
OUT Of
•
0
1
'~
124
142
PARKING
1..-- \i>
I
lANE
lEA\'TNG
<s CZJ
Zl
..!---
I
114
141
VEHICLE
I
IN PARAUEll).NES
I1'HiCLES
l23
PAR~EO
1
I
RlGHTlHT
131
------., ,
122
THRU
'1, ~r
..
MEDI)J(
~-
INTO
PASSENGER AND
MISCELLANEDUS
•
:::~::::
••••••••••••••••
I"'~\:
~~,
140
UTURN
~
lffTREAR
--'L!.
RIGHTTURNING
............. ~.-::-:.~.-:
1-···
..· ···.
................
UfT
,
(,
I
I
U TURN
--..-
'~
104
130
END
~.!.-J
121
RIGHlTHRU
112
113
NEAR
.................
WAWNG
--. ----L
1
,
,
~l~_
OFF PATH
ON CURVE
•
~C2J1
~~
I
RIGHT
.=::::+::::
..
REAR
OFF PATH
ON STRAIGHT
ON PATH
,
I
120
'1 -.-
101
.
ON
(not Ovtl1'k"'ll
DVERTAKING
SAME l).NE
IN
I
SIt(
HEAD
110
TRAffiC
'-r,-
EM[RGING
STANDING
",.Olto
1·0111(1
~::.~~~
.. "::.:::'
'~::J:;
FAR
I·
MANOEUVRING
198
188
178
OTHER
OIHER
OTHER
ON PATH
STRAIGHT
CURVE
?
•
129
139
149
159
169
179
189
UN~NO,/(N
199