Click for next page ( 57

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 56
56 This chapter outlines a proposed plan for such a long-term new investigators may have to be hired and trained and there data collection effort. Unlike the work plan for the current may be requirements for additional office space and equipment. study, this proposed long-term data collection plan is more at However, the setup costs should be only a fraction of the cost the conceptual level. If and when this proposed plan is adopted required to establish a new data collection system. Second, with for implementation, it will then be necessary to develop a more supplemental field data collection, the portion of the CDS cases detailed data collection plan. involving single-vehicle, ran-off-road crashes will be available for use at a relatively low cost. Third, under the NASS special 5.2 Data Collection Alternatives study subsystem, cases may be selected outside of the CDS sam- ple to address specific types of crashes under study. As discussed previously, there were three basic alternatives The proposed long-term data collection plan is, therefore, for the data collection effort in the current study: built around the NASS CDS data collection system, including both within-sample supplemental data collection and outside- 1. New data collection system sample special studies. Note that while NHTSA has maintained 2. Prospective special study under the NASS CDS program the philosophy of allowing the NASS infrastructure to be used 3. Retrospective supplemental data collection for existing for other data collection needs, there are requirements that the NASS CDS cases special study: The retrospective approach is too limited in terms of data Should not have an adverse affect on normal NASS items that can be collected and in flexibility. Some of the desired operations data elements are perishable, i.e., lost after a period of time. For Should not reduce the current NASS CDS caseload for example, data on the struck object would be lost after repair of researchers the object. This information could be necessary to assess the Should not have any impact on current CDS data collec- pre-impact characteristics and conditions of the object as well tion procedures and data elements being collected as to determine its impact performance. The sampling scheme Should not have any impact on NASS operational costs is dictated by the CDS since only sample cases within it are avail- Must cover all costs associated with the development and able. For certain types of crashes, a very long time would be operation of the study required before the sample size becomes sufficiently large for Should be within the interests and expertise of the National proper analysis. Furthermore, NHTSA changed its policy in Center for Statistics and Analysis (NCSA) 2003 so that police accident reports are no longer a part of the Must conform with NHTSA privacy guidelines regarding final NASS case. Police reports are maintained at the Zone Cen- ters for only one year to allow for quality control procedures. collected data Must use existing NASS contractors for all data collection This change in policy will, in essence, eliminate the use of the retrospective approach. and quality control operations Should use a feasibility study to appraise the likely impact The establishment of a new data collection system is a viable, but expensive approach. As discussed previously, there will be and success of the study an initial setup cost for the data collection teams, such as hiring Should use a pilot study in the development of formalized of new personnel, establishing and furnishing the offices, pur- procedures chasing the necessary equipment for conduct of crash investi- Should present to NHTSA an analysis plan, i.e., what re- gation, etc. The investigators will then have to be trained search questions are to be answered extensively in the basics of in-depth level crash investigation, including both classroom and on-the-job training. Then, there These considerations are addressed in the development of is the need to establish cooperation with the local agencies, such the proposed data collection plan presented in the following as law enforcement agencies for the notification system, vehicle section. towing and repair facilities for access to the involved vehicles, hospitals and clinics for medical records/information on injury 5.3 Proposed Data Collection Plan severity, and transportation agencies for highway-related infor- mation. It is also necessary to establish quality control proce- The proposed data collection plan would have two major dures to assure that the data collection effort is conducted subsystems, both of which would be prospective in nature (i.e., properly in terms of validity and accuracy. the cases would be sampled from new crashes): The most efficient and economical approach is to make use of the existing NASS data collection system. First, the initial 1. Continuous sampling subsystem within the CDS sample, setup cost will be greatly reduced since the NASS data teams are and already in place. Depending on the nature of the data collection, 2. Special study subsystem outside the CDS sample.

OCR for page 56
57 The continuous sampling subsystem is intended for a 1. Roadway and traffic characteristics general database to address items of interest pertaining to 2. Roadside characteristics single-vehicle, ran-off-road crashes. This general database 3. Struck-object characteristics would be similar to the database developed under this study. This continuous sampling subsystem would consist For the type of questions that are of interest to the roadside of selecting eligible cases from within the CDS sample and safety community, information on the roadway, traffic, road- supplementing the basic CDS data with additional field side, and struck-object characteristics would be needed for the data on roadway, roadside, and struck-object characteris- analyses. Thus, it would be necessary to collect supplemental tics. In addition, the cases would be reconstructed to the field data on these data elements. Some of the supplemental extent possible to estimate impact conditions and vehicle data, such as highway type, functional class, and traffic charac- trajectories. teristics, would be obtained from the local or state transporta- The special study subsystem would be ad hoc in nature, tion agencies, and cooperation would need to be established intended to address specific questions or roadside safety with these agencies. features. For example, a special study may be designed to Note that even with the supplemental data collection, the assess the impact performance of guardrail terminals. In level of detail on struck-object characteristics would still be lim- order to assure a sufficient sample size to properly assess ited. First, there are simply too many roadside features to the field impact performance, the special study may have to include in the data collection protocol for any details to be col- select cases from outside as well as within the CDS sample. lected on a particular roadside feature. Second, given the intent In addition to the basic CDS data and the supplemental of a general database on single-vehicle, ran-off-road crashes, field data on roadway and roadside characteristics, detailed overly detailed information on struck objects would be overkill. information would be collected on the safety device of Furthermore, it would be very difficult and costly to train the interest. Again, the cases would be reconstructed to the investigators on the details of all these roadside features. The extent possible to estimate impact conditions and vehicle special study subsystem is the more appropriate vehicle for col- trajectories. lecting detailed information on selected roadside features. More detailed descriptions of these two subsystems are It is anticipated that the supplemental field data elements for presented in the following sections. the continuous sampling subsystem would be similar to those used in the current study, with perhaps a few more data ele- ments and additional photographs. It is also anticipated that 5.3.1 Continuous Sampling Subsystem there would be additional coding on information pertaining to As mentioned above, the continuous sampling subsystem impact conditions and vehicle trajectories based on the basic is intended for a general database on single-vehicle, ran-off- CDS data, scaled diagram, and supplemental field data. Finally, road crashes. The cases would be selected from within the the cases would be reconstructed to estimate the impact speeds. NASS CDS sample using sampling criteria similar to those One key consideration is how the supplemental field data used with the retrospective approach in the current study, i.e., would be collected. There are basically two approaches for the data collection: Area type--rural and suburban Single-vehicle, ran-off-road crashes Existing NASS investigators Passenger-type vehicles only--passenger cars and light Newly hired and specially trained investigators trucks Speed limit 75 km/h (45 mph) For the continuous sampling subsystem, the use of existing NASS investigators would be the more logical and cost-effective The sampling criteria may be modified periodically to means of collecting supplemental field data. Based on previous change the range of eligible cases. For example, the area type experience with the retrospective studies, the additional time may be expanded to include urban areas with speed limits required to collect and code the supplemental field data is esti- of 65 km/h (40 mph) and slower, or the vehicle type may be mated to be no more than two hours per case. For a given expanded to include single-unit trucks and tractor-trailers, Primary Sampling Unit (PSU), the number of eligible cases depending on the questions to be addressed with the data. is likely to be less than 50 per year. Thus, the additional time Also, since the cases would be selected within the CDS sample, devoted to the supplemental data collection would not be more the notification system would be the same as the CDS. than 100 hours per year per PSU, or less than two hours per The basic data elements collected under NASS CDS are very week per PSU. It is evident from the estimated workload that it extensive in areas pertaining to the vehicle and occupants, but would not be cost effective to hire an additional investigator per are lacking in detail in the areas of: participating PSU for this supplemental field data collection.