ANALYTICAL SUPPORT FOR PRIORITIZING STATE HIGHWAY PLANS

The following document is a synthesis of multiple project display methods to augment the system of prioritizing state highway plans. These methods provide multiple options to compare and contrast the different project scoring possibilities. The following graphs are based on a Microsoft Excel workbook provided by Chap Tucker, Policy and Planning Manager for the Virginia Department of Transportation, Transportation and Mobility Planning Division. This document is organized in five parts: Project selection by jurisdiction, Comparison of Project Scores, Examination of Range of Project Scores, Comparison of Criteria Weights,

1. Project Selection by Jurisdiction

The following figure gives a visualization of seeing projects across two different dimensions: 1) cost effectiveness, which is not a part of the final project score, and 2) the total project score. Further, the points on the scatter plot are categorized based on the jurisdiction the project serves. This figure is useful because it enables a straight forward way of examining where most projects are serving and how effect the projects are relative to other projects.

Figure 1. Cost Effectiveness v Total Project Score. 14 March 2005.Richmond district characterization of cost effectiveness and the total project score.

2. Comparison of Project Scores:

The purpose of the display method is to graphically illustrate scores for metrics of each goal for each project individually. These radar charts display a footprint of the scoring results. Looking at the radar charts, the user can visually compare the scoring footprints of many individual projects simultaneously. The shape of each graph display how each variable is weighted compared to other products This allows the decision makers to quickly see which projects fit in to which category for example if environmental considerations contribute the most to the score or if level of service has a higher value. They also allow the user to quickly compare one attribute of a project to the same attribute on another project.

Figure 2. Project Score footprint analysis.

Figure 3: Key for Proejct Score Footprint analysis

3. Examination of Range of Project Scores.

The following are the calculations of the mean, min, max, and median of each of the scores for the projects in the Richmond district. This allows the decision maker to compare the value of each variable. These variables allow the decision maker to determine if one of the variables is contributing to the score or not. If there is a low spread in the values this can be seen that this variable is not as significant a variable with a large spread over the projects.

Mean / Maximum / Minimum / Median
2003 LOS / 0.3623 / 0.73 / 0.15 / 0.44
2003 VC / 0.3452 / 0.73 / 0.07 / 0.29
Flow Rate / 0.4495 / 0.73 / 0.07 / 0.44
Access / 0.1095 / 0.73 / 0 / 0
Accident Rate / 0.9385 / 1.38 / 0.14 / 0.97
STRAHNET/ Evacuation / 0.0279 / 0.92 / 0 / 0
Inadequate Geometrics / 0.0000 / 0 / 0 / 0
# Heavy Trucks / 0.5468 / 0.9 / 0.09 / 0.54
Unemployment Rate / 0.0245 / 0.54 / 0 / 0
Environmental issues / 0.5791 / 0.75 / 0.21 / 0.54
Stay in R/W / 0.3636 / 0.75 / 0 / 0
Include HOV, Bike/Ped other modes / 0.3939 / 0.5 / 0 / 0.5
Bridge Sufficiency Rating / 0.0485 / 0.5 / 0 / 0
Cost Effectiveness (cost/vmt) / 0.2455 / 0.5 / 0.05 / 0.25

The following charts show the variance of the scores across all 66 projects and 15 categories.

This chart displays the min, mean, and max of each of the scores in the categories. As you can see the spread on the STRHNET/Evacuation and Inadequate Geometrics have a low spread of in the score showing a low contribution to the decision making process.

Figure 5: This chart displays the max, min and median of the scores and a trend line connection the points. 14 March 2005

Figure 6: This is the same as the first chart but with different methods of displaying the variables spreads.

Figure 7: This chart displays the max, min and mean of the scores without a trend line connection the points. 14 March 2005

Figure 7: This chart displays the max, min and median of the scores without a trend line connection the points. 14 March 2005

4. Comparison of Criteria Weights

This is sensitivity analysis each of the first 15 projects it provides seven different weighting methodologies of the criteria as can be seen in the chart below. This analysis provides the ability to see if changing the weights will change the ranking of the projects. This could be expanded all the project and allow for a sorting of the projects based on the new weights.

Figure 8 This displays the min max and median of the 7 different weighting scenarios for each of these projects.

1 / 2 / 3 / 4 / 5 / 6 / 7
Mobility and Connectivity / 29% / 35% / 60% / 10% / 10% / 47% / 23%
Safety/Security / 23% / 35% / 10% / 60% / 10% / 23% / 47%
Economic Development / 18% / 10% / 10% / 10% / 60% / 10% / 10%
Community and Environmental Preservation / 15% / 10% / 10% / 10% / 10% / 10% / 10%
System Management and Preservation / 15% / 10% / 10% / 10% / 10% / 10% / 10%
Multimodalism / 0% / 0% / 0% / 0% / 0% / 0% / 0%
100% / 100% / 100% / 100% / 100% / 100% / 100%

Table 1: This table lists the 7 different weighting scenarios that were used in the above sensitivity analysis

5. Project Comparison Based on Accident Rate, Flow Rate, and Cost.

GOAL: To enable quick comparison of projects with a subset of metrics