The Investigation Process Research Resource Site
A Pro Bono site with hundreds of resources for Investigation Investigators
Home Page Site Guidance FAQs Old News Site inputs Forums
. . . .. . . . . . . . last updated 8/8/09


to advance the
State-of-the-Art of
investigations, through
investigation process

Research Resources:

Search site for::

Launched Aug 26 1996.


This section attempts to illustrate two points::

  • how a framework for thinking about accidents can influence investigations, and

  • The section contains papers based on observations by investigators during investigations that were not necessarily included in the investigaiton reports,

The section contains papers ranging from in-house research study at the NTSB to develop a framework for thinking about hazamat transportation safety and its regulation, to findings based on observations made by investigators over several investigations in different modes that did not find their way into the investigation reports, primarily because of thei report's focus on determination of "cause" or "probable cause."

The effects of the Risk Analysis Concepts study provided a framework for investigating hazmat accidents, fires, explosions and releases from late 1971 until at least 1982, with noteworthy results. The consolidation of observations and other reported data, and the temporal and spatial ordering of the data enabled the systhesis of the behavioral process for released hazardous materials and responders in emergencies. As so often happens with serendipitous discoveries, this alternative research method and its significance remained unrecognized until years later.

More about this issue will be posted here later.

  1. Risk Concepts Special Study Following Star's seminal puiblication of "An Overview of the Problem of Public Safety" the NTSB supported an early special study titled Risk Conccepts in Dangerous Goods Transportation,/ The study was a form of internal research by the Board into safety concepts in the hazardous materials transportation field. This research report proposed a risk-based framework for hazmat safety regulation. The significance from the perspective investigation process research is that it provided a framework for subsequent hazmat-related investigations conducted at the NTSB from 1970 to 1982. It also was one of the outputs that stimulated discussion and research projects that explored risk analyes of hazardous materials transportaiton activities, including a Conference sponsored by a transportation umbrella group. Proceedings of a Conference about the Risk Concepts Study. The conference elaborated on the proposed framework and helpded advanced interest in the approach, and receptiveness to risk studies. Research studies addressed hazmat routing, terminal siting, regulation, hazmat release dispersion and other risk.

  2. The Story of GEBMO and DECIDE

    An understanding of hazmat behavior processes in incidents and of responder decision processes during emergencies was developed from direct observations during investigaitons of incidents in several modes, and data from other public sources about the incidents. The models were derived from the obsrvation and ordering of the observed actions during investigaitons, and then systheses of models of generalize descriptions of hazardous materials release and emergency decision "processes" and subsequent "validation" by observed behaviors in subsequent incidents were a derivative of the scientific method. The models helped create a new paradigm for emergency responses to hazardous materials accidents.

    • List of accident cases investigated by NTSB 1968-1982 referenced in the paper describing how the hazmat GEBMO and DECIDE models were developed. Includes synopsis of listed cases.

    • Lesson Book

      Created for hazardous materials emergency responder junior college course, and developed during the investigations of the accidents in the case list. List of references available at that time is also provided. In this case, the research data supporting the course was derived from direct observations during investigations, sequentailly ordered and linked to show coupled reactions, in the manner of somewhat similar to Gantt chart, first published in a NTSB's report of a 1970 motor vehicle liquid oxygen explosion in Brooklyn, NY, described in this paper. This process, with initial refinement, was subsequently documented in a 1975 paper. Journal of Safety Research report.

      The lesson book was the first to provide categories of hazmat dispersions, aggregated from observations during investigated accidents and other accident reports. Subsequently, modeling of such behaviors with computer models has become much more sophisticated, but the general patterns remain valid, and for the basis for emergency training to this day.

  3. Example of institutional issues

    The standards development issue evolved from the application of Risk Concepts framework and MES tmethods in cases involving emergency responder casualties. By probing the successive steps that led to the losses well prior to the injuries disclosed deficiencies in the process by which the standards for behaviors in such situations were promulgaged. These training problems were described in paper titled Safety Training: Safety's Achilles Heel" Another example of actionns that influenced behaviors subsequent to investigations, based on observations during investigations, addresss the coupling of predictive safety analyses with verfication of those predictions as part of the investigation process. System Safey's Open Loop.

  4. The Marjorie McCalister investigation.

    The NTSB investigated a case involving the sinking of a vessel of f the US coast in bad weather. The vessel was not recovered. In the absence of physical data, the Board 's intermodal staff used logic trees based on survivor's information and vessel design data for the first time to develop alternative hypotheses to explain the sinking. This was in effect a research effort to develop a way to organize data acquired during investigation to help arrive at potential scenarios that might explain what happened. The tree was published in the Board's report to illustrate the Board's handling of the investigtaion data. When the vessel was raised after the report had been issued, the most likely reason for the accident was confirmed by the condition in which the boat was found.

    The intent for publishing the tree was to describe the way the speculation about the accident scenario was organized and how the scant data available was used to arrive at the most likely scenario, introducing the use of the method to investigation processes. It was not intended to be used as a method of choice for describing and explaining the sinking , because it could not be verified with surviving data, had gaps which could not be linked, and did not present time relationships among the entries.

    (More to follow. . . )