Page 113

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

APPENDIX B

Survey Responses

Contact Information

2. What state DOT do you represent?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Open ended	46	100	AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, SC, SD, TN, UT, VT, VA, WA, WI, WY

Use of SPFs

5. Does your agency use SPFs included in the Highway Safety Manual or developed by another agency?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Yes	33	72	AZ, CA, DE, FL, GA, ID, IL, IN, KS, KY, ME, MD, MI, MN, MS, MO, NV, NJ, NY, NC, OH, OK, OR, PA, SC, SD, TN, UT, VT, VA, WS, WI, WY
No	13	18	AL, AK, AR, CO, CT, IA, LA, MA, MT, NE, NH, NM,ND

SPF Calibration

6. Does your agency apply calibration factors to these SPFs?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Yes	22	67	FL, ID, IL, KS, ME, MI, MN, MS, MO, NJ, NY, NC, OH, OR, PA, SC, TN, UT, VT, VA, WI, WY
No	11	33	AZ, CA, DE, GA, IN, KY, MD, NV, OK, SD, WS

Note: In the following tables, a blank space after the colon following a state abbreviation means that there was no comment from that state.

8. For what facility types does your agency have calibration factors developed? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Design-level: Two-lane rural roadway segments	20	95	FL, ID, IL, KS, ME, MN, MS, MO, NJ, NY, NC, OH, OR, PA, SC, TN, UT, VT, WI, WY
Design-level: Multi-lane rural roadway segments	16	76	FL, IL, KS, MN, MS, MO, NJ, NY, NC, OH, OR, PA, SC, TN, UT, WI
Design-level: Urban-suburban arterial roadway segments	15	71	FL, IL, ME, MS, MO, NJ, NY, NC, OH, OR, PA, SC, TN, UT, WI
Design-level: Basic freeway segments	11	52	IL, KS, MO, NY, NC, OH, PA, SC, TN, UT, WI

Page 114

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

Design-level: Freeway segments with speed change lanes	5	24	KS, NY, OH, PA, TN
Design-level: Intersections on two-lane rural roads	18	86	FL, ID, IL, ME, MN, MS, MO, NJ, NY, NC, OH, OR, PA, SC, TN, UT, VT, WI
Design-level: Intersections on multilane rural roads	15	71	FL, IL, MN, MS, MO, NJ, NY, NC, OH, OR, PA, SC, TN, UT, WI
Design-level: Intersections on urban-suburban arterials	16	76	FL, IL, KS, ME, MS, MO, NJ, NY, NC, OH, OR, PA, SC, TN, UT, WI
Design-level: Freeway ramps	5	24	KS, NY, OH, PA, TN
Design-level: Other	6	29	KS: Crash Modification Factors; ME: Roundabouts; NY: ; OH: Roundabouts, Ramp Terminal Intersection Rural Signalized D4, Ramp Terminal Rural Unsignalized D4, Ramp Terminal Urban Signalized A2, A4, B2, B4, D3EX, D4, and Ramp Terminal Urban Unsignalized A2, B2, B4, D3EX, D4; PA: Roundabouts; UT: Segments only consider CMFs that were readily available
Network-screening-level: Two-lane rural roadway segments	5	24	ID, KS, MI, MS, NY
Network-screening-level: Multi-lane rural roadway segments	4	19	KS, MI, MS, NY
Network-screening-level: Urban-suburban arterial roadway segments	4	19	KS, MI, MS, NY
Network-screening-level: Basic freeway segments	3	14	KS, MI, NY
Network-screening-level: Freeway segments with speed change lanes	1	5	NY
Network-screening-level: Intersections on two-lane rural roads	4	19	KS, MI, MS, NY
Network-screening-level: Intersections on multi-lane rural roads	4	19	KS, MI, MS, NY
Network-screening-level: Intersections on urban-suburban arterials	4	19	KS, MI, MS, NY
Network-screening-level: Freeway ramps	1	5	NY
Network-screening-level: Other	2	10	NY: ; PA: Roundabouts

9. Who developed these calibration factors? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Someone within your agency	3	14	ME, MI, OH
External Consultant	13	62	IL, KS, MI, MN, MS, MO, NJ, NY, OH, PA, UT, WI, WY
Academic Researcher	12	57	FL, ID, KS, MI, MO, NC, OH, OR, PA, SC, TN, VT
Another state DOT or local agency	0	0

Page 115

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

10. What was the approximate cost required to estimate these calibration factors?

Answer Choices	Responses	Percentage of Responses	Individual Responses
< $10,000 per facility type	1	5	NY
$10,000 - $25,000 per facility type	3	14	MO, OH, VT
$25,000 - $50,000 per facility type	2	10	MI, TN
> $50,000 per facility type (please elaborate)	4	19	KS: Research projects may have included 1-4 facility types at $50-100K each; MN: ; PA: Calibration of the Freeways, Ramp segments, speed change lanes, and ramp terminals cost PennDOT $388K using the consultants Kittelson & GTE. The roundabouts calibration of Report 888 cost PennDOT $58K using the consultant Kittelson. The regionalized SPFs developed by Penn State Univ. also have county level calibration factors. I can't determine how much the county calibrations cost separate from the SPF development; UT:
			FL: > $100,000 in total; ID: ; IL: ; ME: ; MS: ; NJ: ; NC: Around $15,000 per facility type, however, my answer is given assuming that “facility type” includes all segments and intersections in that facility type as well as all the regional calibration factors that are developed (we have three regions in NC and develop calibration factors for each one); OR: ; SC: ; WI: Hard to determine a cost per calibration factor as a lot of other work went into the projects and multiple projects were used. The most expensive part was
Unsure	11	52	the data collection; WY:

11. What types of tools were used to develop these calibration factors? Select all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Manual Calculations	6	29	MN, MS, MO, NY, OH, SC
Spreadsheet	15	71	FL, IL, ME, MN, MS, MO, NJ, NC, OH, OR, PA, SC, TN, UT, WY
FHWA Calibrator Tool	4	19	ME, MS, NJ, WI
Other (please elaborate)	5	24	KS: Research study by University. See attachments; MI: SafetyAnalyst; NJ: All in the FHWA-NJ-2019_007 Report; VT: SPSS; WI: GIS and other applications
Unsure	1	5	ID

12. What type of calibration factor does your agency typically use?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Calibration factor defined in the HSM (single value equal to the sum of observed crash frequency divided by sum of predicted crash frequency)	15	71	FL, ID, IL, ME, MN, MO, NJ, NC, OH, OR, PA, SC, TN, UT, VT
Calibration function	1	5	MI

Page 116

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

Other (please elaborate)

3

14

KS: Both; MS: Both; WI: We use a calibration factor as defined in the HSM, but also developed functions for the day when IHSDM (or another tool) can utilize them.

13. Approximately how often are these calibration factors updated?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Annually	1	5	VA
Every 2-3 years	2	9	UT, WY
Every 4-5 years	2	9	NY, NC
More than every 5 years	4	18	KS, MO, PA, WI
Never, they were only estimated once	11	50	FL, ID, ME, MI, MN, MS, OH, OR, SC, TN, VT
Unsure	2	9	IL, NJ

14. Does your agency estimate multiple calibration factors for a given facility type to account for regional differences acro ss your agency?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Yes	5	24	IL, NY, NC, SC, TN
No	12	57	FL, ID, KS, ME, MI, MS, MO, NJ, OH, OR, UT, WY
Sometimes (please elaborate)	4	19	MN: We have looked at this but found little difference across broad geographical areas; PA: Penn State will develop new calibration factors for CMF research. So they did this for the Adaptive Traffic Signals CMF development. We also had updates for our second network screening in 2019/2020. They also developed a special calibration factor for a specific roadway in Philadelphia called the Roosevelt Blvd; VT: A three-region classification was evaluated in the 2019 calibration project: Northern (Grand Isle, Franklin, Lamoille, Orleans, Caledonia, and Essex Counties), Central (Addison, Chittenden, Washington, Orange, Rutland, and Windsor Counties), Southern (Bennington and Windham Counties). There is little variation, and statewide CFs are preferred; WI: No. But this is something we would like to do in the future, assuming we have the data necessary.

15. Why were calibration factors chosen as the preferred method of adjusting to local conditions? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Data availability	9	43	ID, ME, MN, MS, NC, PA, SC, UT, WI
Available staff resources	7	33	ID, ME, MO, NC, PA, SC, UT
Available staff skillsets	5	24	ME, MO, PA, SC, TN
Available budget	8	38	ME, MN, MO, NC, OR, PA, SC, WI
Provides the best value for needed precision to support agency decision-making	12	57	IL, MI, MN, MS, MO, NC, OH, PA, UT, VT, WI, WY
So we do not have to develop our own adjustment factors	2	10	KS, WI

Page 117

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

Other (please elaborate)

5

24

FL: Calibration factors were developed around 2011/2012 after the HSM 1st edition was released. The Department has currently invested to develop FL-specific SPFs for different facilities; NJ: For easier use of HSM tools like Part C spreadsheets where a calibration factor is entered as input variable; NY: Analysis showed many of the facility type SPFs were accurate for NYS data as is but others were adjusted where NYS showed statistical differences; UT: We developed calibration factors as a “gap fill” measure until Network Screening SPFs could be developed by a local University. This effort is used in our Intersection Capacity Evaluations, as Systemic Improvements prioritization efforts (story maps), and our evaluation of design exceptions etc.; WI: WisDOT chose to use IHSDM as our analysis tool. Calibration factors work with this tool and it was free for locals to use as well. This may shift now that IHSDM is no longer updated and has planned obsolescence. Easy to defend since it is included in the HSM.

16. Are the resulting crash frequency estimates obtained using the calibration factors validated in some way? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
No validation or comparison between the predicted crash frequencies and reported crash frequencies is performed	6	29	ID, KS, MO, OR, PA, VT
We compare predictions with reported values for different facility types	10	48	ME, MI, MN, MS, NJ, NC, OH, PA, TN, UT
We compare predictions between uncalibrated and calibrated models	3	14	OH, PA, SC
Examine if calibration factors are within a reasonable range (please elaborate)	5	24	MN: We checked these on cure-plots to validate if the calibration factors worked. This gave us help in deciding if SPF development would be needed; NC: We look to see if any calibration factors are oddly high or low; PA: CURE plots were completed by the consultant; UT: When calibration factors become higher than 2 or 3 we review the data inputs and evaluate if the data provided to the models are prone to error, if so we opt not to publish calibration factors and direct engineers to use a calibration factor of 1 until better information is available. If we feel that there is reason to justify using higher calibration numbers (due to our operational methods) we publish them for use; WI: Metrics are described in the attached reports.
Other (please elaborate)	3	14	IL: Unsure; NY: The majority accurately reflect the average conditions experienced but we are working on improving some that were affected by data errors; WY:

17. How well do the calibrated SPFs represent safety conditions experienced by your agency?

Answer Choices	Individual Responses
Open Response	IL: Unsure; KS: Kansas DOT has recently created our own state-specific SPFs - calibration factors were not used in this development for network screening; ME: Difficulty with some facility types because we had to use our whole network to get enough crashes; MI: as developed the SPFs did represent what was

Page 118

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

present, but we do see the need to update/adjust to reflect current conditions; MO: It depends on the facility. There are some SPFs such as the urban signalized intersection where the SPF predicted crashes require high calibration factor to reflect our state's conditions. We recognize this as an opportunity to develop a state specific SPF and are looking at third party tools to help develop those; NJ: Most checking are only done for ongoing projects, which are usually hot spot locations and predicted crashes appear to be in the same range with occurring crashes; NY: The majority accurately reflect the average conditions experienced but we are working on improving some that were affected by data errors; NC: Generally well, though in many cases, the predicted value is not directly comparable to the observed crash history due to unique circumstances; OH: Historically sufficient, reviewing newer ones, developed our own Freeway SPFs because we weren’t seeing the representation that facility type; OR: Not widely used in engineering practices so not too much information to share; PA: For Freeways, SC lanes, Ramp Segments, and ramp terminals they are pretty good. Roundabouts seem fair; SC: Typically the calibrated values have been reasonable when reviewed based on field/observed data; TN: See attached reports; UT: Generally speaking I think they work well in predicting the total crashes for the given intersections and segments. The FI and PDO values are not as reliable. There are some locations where the observed frequencies are higher than the calibrated predicted models and some where the values are lower than the calibrated predicted models. We use the 5-year severity distribution values by facility type rather than severity distribution functions. This is to simplify the process. Keeping in mind that individual severity by collision types vary drastically require some level of judgement, for example in our ICE program we identify problematic areas only when the observed crash types and severities and predicted crashes are >50% higher than predicted values being sensitive to overlook observations that only consider one crash; VT: There is a feeling that they may predict lower than actual crash frequency. We are hoping to recalibrate them in the near future. WI: Started using SPFs in 2019. Projects analyzed are just being built now, so don't have good observed data yet

18. What are the biggest challenges to estimating calibration factors for your agency? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Data Availability	10	48	ID, IL, MS, NJ, NC, OH, UT, VT, WI, WY
Available staff resources	13	62	ID, IL, KS, ME, MI, MN, MS, MO, NJ, NY, PA, SC, UT
Available staff skillsets	14	67	FL, ID, IL, KS, MI, MN, MS, NJ, PA, SC, TN, UT, VT, WI
Available budget	10	48	IL, ME, MI, MO, OH, OR, PA, SC, VT, WI
Other (please elaborate)	3	14	OH: Having enough site types; PA: Combining data sets from all of the different PennDOT divisions and ensuring it is recent. Also, we lack complete information on our locally owned roads. So, it is pretty tough to complete Calibrations for SPFs on locally owned roads; WI: Having reliable, accessible, and easy to use data.

19. You indicated that you used a calibration factor developed by another agency. How does your agency determine if the calibration factor accurately reflects conditions experienced by your agency? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
We don’t	0
Compare predictions for different facility types	0
Compare predictions from uncalibrated and calibrated models	0
Check if calibration factors are within a reasonable range (please elaborate)	0
Other (please elaborate	0

Page 119

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

Non-use of SPF calibration

20. You indicated that your agency does not estimate or apply calibration factors to externally developed SPFs. Why does your agency not estimate or apply calibration factors? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Data Availability	2	18	AZ, OK
Available staff resources	5	45	CA, DE, NV, SD, WS
Available staff skillsets	1	9	CA
Available budget	2	18	AZ, DE
Uncalibrated SPFs provide precision needed to support agency decision-making	4	36	CA, KY, MD, SD
Other (please elaborate)	7	64	CA: Originally, due to California data being used in some way in the development of many of the HSM models, Caltrans is of the position that the HSM results currently adequately capture the California experience. However, Caltrans is beginning to undertake HSM model calibration. A “framework” has been developed which informs the data needs, person-hours, and approximation of costs to calibrate the HSM for Caltrans. Further, Caltrans safety analysis priorities have also been included which informs how the model calibration prioritization has been developed. Caltrans will likely be moving forward with actual calibration efforts in the near future, informed by the aforementioned framework. The discussion of calibrating the HSM1 vs HSM2 models is TBD, and will further inform future calibration efforts. We can send over our framework. Once it's ADA-remediated, it will be uploaded to Caltrans' external HSM website: https://dot.ca.gov/programs/design/manual-highway-safety-manual-hsm ; DE: We are currently in the process of developing Delaware-specific calibration factors; GA: AASHTOWare Safety/Numetric uses our own data to create SPFs so we do not need to calibrate them since they are already made with our own data; IN: We developed our own SPFs using Indiana data. Do not use externally developed SPFs; IN: We developed our own SPFs using Indiana data. Do not use externally developed SPFs; MD: We developed statewide calibration factors about ten years ago. The study is outdated and the sample size did not meet minimum AASHTO requirements for many facility types; NV: Knowledge gaps; WS: Many of the HSM SPFs were developed using WA state data.

Use of agency-specific SPFs

21. Has your agency used agency-specific SPFs developed specifically for your agency .

Answer Choices	Responses	Percentage of Responses	Individual Responses
Yes	30	65	AL, AZ, AR, CO, CT, FL, GA, ID, IL, IN, IA, KS, KY, LA, MA, MI, MN, MT, NV, NY, OH, OK, PA, SC, SD, TN, VA, WS, WI, WY
No	16	35	AK, CA, DE, ME, MD, MS, MO, NE, NH, NJ, NM, NC, ND, OR, UT, VT

Page 120

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

Agency-specific SPFs

23. For what facility types does your agency have agency-specific SPFs? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Design-level: Two-lane rural roadway segments	10	45	CO, CT, GA, ID, IL, IN, NV, PA, SC, TN
Design-level: Multi-lane rural roadway segments	8	36	CO, CT, GA, IL, IN, PA, SC, TN
Design-level: Urban-suburban arterial roadway segments	8	36	CT, GA, IL, IN, PA, SC, TN, WS
Design-level: Basic freeway segments	8	36	CO, CT, GA, IL, IN, OH, PA, TN
Design-level: Freeway segments with speed change lanes	4	18	CT, IN, OH, TN
Design-level: Intersections on two-lane rural roads	10	45	CO, CT, GA, ID, IL, IN, KS, PA, SC, TN
Design-level: Intersections on multi-lane rural roads	9	41	CO, CT, GA, IL, IN, KS, PA, SC, TN
Design-level: Intersections on urban-suburban arterials	10	45	CO, CT, GA, IL, IN, KS, MA, PA, SC, TN
Design-level: Freeway ramps	3	14	CT, IN, TN
Design-level: Other	6	27	AZ: Roundabouts; FL: Signalized intersections by Florida Context Class; GA: ; MN: Specific Treatment (HTCB); PA: Urban/Suburban Collector Roads; TN: 5-lane rural roads
Network-screening-level: Two-lane rural roadway segments	16	73	AL, AZ, CO, CT, GA, IN, IA, KS, KY, LA, MA, MI, MT, OK, SD, WI
Network-screening-level: Multi-lane rural roadway segments	15	68	AL, AZ, CO, CT, GA, IN, KS, KY, LA, MA, MI, MT, OK, SD, WI
Network-screening-level: Urban-suburban arterial roadway segments	13	59	AL, CT, GA, IN, IA, KS, KY, LA, MA, MI, OK, SD, WI
Network-screening-level: Basic freeway segments	13	59	AL, CO, CT, GA, IN, IA, KS, KY, LA, MI, MT, SD, WI
Network-screening-level: Freeway segments with speed change lanes	2	9	CT, IN
Network-screening-level: Intersections on two-lane rural roads	13	59	CO, CT, GA, IN, IA, KS, KY, LA, MI, MT, OK, SD, WI
Network-screening-level: Intersections on multilane rural roads	13	59	CO, CT, GA, IN, IA, KS, KY, LA, MI, MT, OK, SD, WI
Network-screening-level: Intersections on urban-suburban arterials	14	64	AL, CO, CT, GA, IN, IA, KS, KY, LA, MI, MT, OK, SD, WI
Network-screening-level: Freeway ramps	3	14	CT, IN, TN
Network-screening-level: Other	4	18	FL: Segments and signalized intersections based on Florida Context Class; GA: ; KY: Roadway departure on rural 2-lane roads, median crossover multilane; LA: Roadway departure, pedestrians

Page 121

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

24. Who developed these SPFs? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Someone within your agency	5	19	AZ, KY, LA, MI, OK
External consultant	17	63	AL, AZ, AR, CO, FL, GA, IL, KS, LA, MA, MI, MN, MT, OH, SD, TN, WI
Academic researcher	13	48	AR, ID, IN, IA, LA, MA, MI, NV, OH, PA, SC, TN, WS
SPF was developed by another agency (please elaborate	1	4	CT: Contracted team affiliated with University of Connecticut

25. What is the approximate level of effort required to estimate these SPFs?

Answer Choices	Responses	Percentage of Responses	Individual Responses
< $20,000 per SPF	10	40	AR, CT, IN, KS, KY, LA, MA, MT, OH, SD
$20,000 - $40,000 per SPF	4	16	CO, FL, MI, OK
$40,000 - $60,000 per SPF	4	16	AL, GA, IA, TN
$60,000 - $80,000 per SPF	0	0
$80,000 - $100,000 per SPF	3	12	AZ, MN, WS
> $100,000 per SPF (please elaborate)	5	20	AL: ; IL: Cost is unknown; NV: Unknown; PA: From what I can gather these SPFs cost over $1 million to develop. The additional SPFs for urban/suburban collector roads was about $90K; WI: Hard to say what the actual cost per facility type is. Definitely lower than $100,000 per. But was part of the same projects as the Calibration Factors.

26. Approximately how often are these SPFs generally updated?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Annually	4	15	FL, KS, KY, TN
Every 2-3 years	3	11	IN, IA, MA
Every 4-5 years	6	22	AR, CO, CT, IL, MT, SD
More than every 5 years	3	11	OK, PA, WI
Never, they were only estimated once	9	33	AL, ID, LA, MI, MN, NV, OH, SC, WS
Unsure	2	7	AZ, GA

27. Does your agency estimate multiple SPFs or somehow adjust SPF predictions for a given facility type to account for regional differences across your agency?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Yes	7	26	AR, FL, GA, MN, OK, PA, WS
No	19	70	AL, AZ, CO, CT, ID, IN, IA, KS, KY, LA, MA, MI, MT, NV, OH, SC, SD, TN, WI
Unsure	1	4	IL

28. Why were agency-specific SPFs chosen as the preferred method for accounting for local conditions? Check all that apply.

Answer Choices

Responses

Percentage of Responses

Individual Responses

Page 122

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

Precision needed for crash frequency predictions	13	57	AZ, CO, CT, FL, IL, IN, IA, MI, MN, OH, OK, PA, SC
Do not have specific data/ variables to apply to other SPFs	3	13	FL, OK, PA
Needed to capture regional differences across the state	5	22	AR, GA, IL, OK, PA
Had sufficient in-house expertise to estimate SPFs	7	30	AZ, CO, CT, GA, IN, KY, TN
Had sufficient resources available to support project to develop SPFs	11	48	AZ, CO, CT, FL, IN, IA, KS, LA, MA, PA, WS
Other (please elaborate)	6	26	AR: SPF developed by consultant/academic researcher as part of developing Arkansas safety knowledge base in concert with deployment of HSM compliant software Vision Zero Suite; KS: the opportunity presented itself to develop state-specific SPFs. We still lack sufficient data to do regional analysis; MA: See attached report, it was a better fit; MT: I do not have the precise reason - a decision made by our Traffic & Safety predecessors; PA: Pennsylvania has many regional differences. So, a statewide SPF or calibration factor in many cases would not be the best option; WI: WisDOT only developed network screening level SPFs. These are needed to perform the Level of Service of Safety (LOSS) and Potential for Safety Improvement (PSI) metrics which are used to determine locations of interest

29. Does your agency combine adjustment factors from SPFs in the HSM with SPFs developed by your agency?

Answer Choices	Responses	Percentage of Responses	Individual Responses
Yes (please elaborate)	3	11	CO: CMFs from the HSM are used in screening and predictive analysis, but not necessarily as an adjustment for the SPFs, but rather in economic analysis; KY: When available; MA: Some
No	21	78	AL, AZ, CT, FL, GA, ID, IN, IA, KS, LA, MI, MN, NV, OH, OK, PA, SC, SD, TN, WS, WI
Unsure	3	11	AR, IL, MT

Non-use of agency-specific SPFs

30. Why does your agency not estimate its own SPFs to reflect agency -specific conditions? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Data Availability	7	47	AK, MD, NE, NJ, NC, ND, OR
Available staff resources	10	67	AK, DE, ME, MD, MS, MO, NE, NH, NJ, ND
Available staff skillsets	5	33	MD, NE, NH, NJ, ND
Available budget	5	33	DE, ME, MD, MO, OR
Other methods provide precision needed to support agency decision-making	4	27	AK, CA, NC, ND
Other (please elaborate)	3	20	NJ: With calibration factors, it is easier to use the HSM tools in conducting crash frequency calculations; NM: To the best of my knowledge, we do have not incorporated any SPFs or safety software into our projects yet. We recently attended a IHSDM training class; UT: We are currently developing network screening specific SPFs with a local university, but their results are not completed at the time of this survey.

Page 123

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

31. What are the biggest challenges to estimating SPFs for your agency? Check all that apply.

Answer Choices	Responses	Percentage of Responses	Individual Responses
Data availability	24	60	AK, AZ, AR, CA, CO, FL, GA, ID, IL, IN, KS, KY, LA, MD, MN, NJ, NC, ND, OH, OK, OR, PA, WS, WI
Available staff resources	26	65	AK, AZ, CA, CO, DE, GA, ID, IL, KS, LA, MD, MA, MI, MN, MS, MO, NV, NH, NJ, ND, OH, OK, PA, SC, SD, WS
Available staff skillsets	23	58	CA, FL, GA, ID, IL, KS, ME, MD, MA, MI, MN, MT, NE, NV, NH, NJ, NC, ND, OH, PA, SC, TN, WI
Available budget	16	40	AZ, CA, CO, DE, FL, IL, MD, MI, MO, NC, OH, OR, PA, SC, WS, WI
Other (please elaborate)	11	28	AR: Quality; CA: Originally, due to California data being used in some way in the development of many of the HSM models, Caltrans is of the position that the HSM results currently adequately capture the California experience. Therefore, CA-specific SPFs are not necessary. Upon investigating HSM model calibration, it was determined that there’s adequate data to develop CA-specific SPFs for some facility types which could be explored in lieu of calibration. Despite the available data and potential incremental increase in precision of utilizing a CA-SPF versus calibration, the resource needs to develop calibration are more favorable than developing SPFs; IA: Convincing the upper management on the robustness of the process; KY: Available AFs; MA: Data quality of crash location; NH: Due to the checked boxes above (staff resources and staff skillsets), it is difficult to know what the challenges are. We are just starting to get into socializing the HSM in our agency; NJ: Aside from the challenges, it is easier to use HSM tools that and readily available; NM: Getting contracts developed. We also have a lot of data, but it is siloed in many different Divisions. It is also in several different formats. So, combining data sets can be a huge lift. We also have to ensure that datasets are current and still relevant; UT: Our biggest challenge is bridging the gap between academia and practice. Often the “research” groups are too focused on specifics rather than understanding the needs of the practitioner; WI: Same data issues as the calibration factors.

Page 124

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

This page intentionally left blank.

Page 125

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.

Abbreviations and acronyms used without definitions in TRB publications:

A4A	Airlines for America
AAAE	American Association of Airport Executives
AASHO	American Association of State Highway Officials
AASHTO	American Association of State Highway and Transportation Officials
ACI–NA	Airports Council International–North America
ACRP	Airport Cooperative Research Program
ADA	Americans with Disabilities Act
APTA	American Public Transportation Association
ASCE	American Society of Civil Engineers
ASME	American Society of Mechanical Engineers
ASTM	American Society for Testing and Materials
ATA	American Trucking Associations
CTAA	Community Transportation Association of America
CTBSSP	Commercial Truck and Bus Safety Synthesis Program
DHS	Department of Homeland Security
DOE	Department of Energy
EPA	Environmental Protection Agency
FAA	Federal Aviation Administration
FAST	Fixing America’s Surface Transportation Act (2015)
FHWA	Federal Highway Administration
FMCSA	Federal Motor Carrier Safety Administration
FRA	Federal Railroad Administration
FTA	Federal Transit Administration
GHSA	Governors Highway Safety Association
HMCRP	Hazardous Materials Cooperative Research Program
IEEE	Institute of Electrical and Electronics Engineers
ISTEA	Intermodal Surface Transportation Efficiency Act of 1991
ITE	Institute of Transportation Engineers
MAP-21	Moving Ahead for Progress in the 21st Century Act (2012)
NASA	National Aeronautics and Space Administration
NASAO	National Association of State Aviation Officials
NCFRP	National Cooperative Freight Research Program
NCHRP	National Cooperative Highway Research Program
NHTSA	National Highway Traffic Safety Administration
NTSB	National Transportation Safety Board
PHMSA	Pipeline and Hazardous Materials Safety Administration
RITA	Research and Innovative Technology Administration
SAE	Society of Automotive Engineers
SAFETEA-LU	Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (2005)
TCRP	Transit Cooperative Research Program
TEA-21	Transportation Equity Act for the 21st Century (1998)
TRB	Transportation Research Board
TSA	Transportation Security Administration
U.S. DOT	United States Department of Transportation

Page 126

Suggested Citation: "Appendix B: Survey Responses." National Academies of Sciences, Engineering, and Medicine. 2024. Calibration and Development of State-DOT-Specific Safety Performance Functions. Washington, DC: The National Academies Press. doi: 10.17226/27906.