Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
E-1  Contents E-1 Introduction E-1 Using the Tool E-2 Example Images Introduction The CMF Combination Tool is a Microsoft Excel spreadsheet implementation of the homo- geneity testing procedure developed under NCHRP Project 17-63: Guidelines for the Develop- ment and Application of Crash Modification Factors. This tool can be used to test two or more CMFs for the same treatment to determine whether they are similar enough to be combined into a single CMF value. The CMFs must also address the same crash type (e.g., total crashes, rear end crashes). If the CMFs are able to be combined, the tool produces a value for the combined CMF. It also gives cautions for the use of the combined value. This tool is not intended for combining CMFs for different treatments. See the Examples below. ⢠Example 1. You have two CMFs that estimate the effect on total crashes of installing wider edgelines. You want to know if they are similar enough to be combined into a single value to be used for an upcoming project to widen edgelines. This tool is intended for this use. ⢠Example 2. You have two CMFs, one for the effect of wider edgelines and one for the effect of rumble strips. You want to calculate a single combined CMF that will provide an estimate of the effect on total crashes if both treatments were installed together. This tool is not for this use. However, guidelines for this purpose are provided under NCHRP Project 17-63. Using the Tool To use the tool, the analyst inputs two or more CMFs and their standard errors (both required) and receives a calculated combined CMF. There must be at least two CMFs entered, but no more than 100 CMFs. Figure E1 shows the input screen from the Main tab of the tool. The top middle portion of the sheet is used to display the resulting combined CMF and caution messages. The top right portion of the screen is used to display a summary of the CMFs entered by the user, including the number of CMFs, the range of the CMF values, and the range of the standard errors. The colored area below these areas is for user input. A P P E N D I X E User Guide for the CMF Combination Tool Spreadsheet
E-2 Guidelines for the Development and Application of Crash Modification Factors The caution messages inform the user about certain conditions that might apply to the results. ⢠Caution 1 alerts the user if the combined CMF is greater than 1.0, indicating that the treat- ment may cause an increase in crashes. ⢠Caution 2 alerts the user if the combined CMF has a standard error that is large enough to put the reliability of the result in question. ⢠Caution 3 alerts the user if more than half of the CMFs have a statistical weight less than 4.0, which calls into question the reliability of the resulting combined CMF. Example Images The screenshots below provide a series of examples showing the tool being used to compare and potentially combine two CMFs. Figure E2 is an example of two CMFs being combined suc- cessfully. Figure E3 through Figure E6 show examples of CMFs that should not be combined or for which the combined result should be used with caution. Figure E1. Main screen of CMF Combination Tool. Figure E2. Example of two CMFs that were combined into a single value.
User Guide for the CMF Combination Tool Spreadsheet E-3Â Â Figure E3. Example of two CMFs that should not be combined because the statistical variability was too large. Figure E4. Example of two CMFs that have a combined result that indicates that the treatment might increase crashes.
E-4 Guidelines for the Development and Application of Crash Modification Factors Figure E5. Example of two CMFs that have a combined result that should be used with caution for producing quantitative estimates of crash frequency reduction because the result may be statistically unreliable due to a large standard error. Figure E6. Example of four CMFs that have a combined result that should be used with caution for producing quantitative estimates of crash frequency reduction because the result may be statistically unreliable due to low statistical weights.