Measuring the Impact of Cognitive Bias

Measuring the Impact of Cognitive Bias in Fire Investigation

Paul Bieber, CFEI, B.S., M.L.S.
Director of the Arson Research Project, USA

Presented at the International Symposium on Fire Investigation Science and Technology, 2012

ABSTRACT

Cognitive bias has been found to shape decision making in a wide variety of fields. Criminal investigation and the forensic sciences are no exception.  Fire investigation, part criminal investigation, part forensic examination, is uniquely positioned to be influenced by the affects of cognitive bias.

The 2009 report from the National Academy of Science, Strengthening Forensic Science in the United States; A Path Forward (NAS Report) , recognizes conceptual bias as a factor in all forensic disciplines.  The National Fire Protection Association Guide for Fire and Explosion Investigation (NFPA 921)  acknowledges these biases as a concern in fire investigation.3

This report will explore the most common forms of cognitive bias found in the field of fire investigation, review past research and give recommendations on how these biases might be minimized. It will also present the results of new research which sought to measure the influence of expectation and role bias in fire investigation. A companion report, “Case Study Review of Contextual Bias in Fire Investigation” is available at www.Thearsonproject.org.

Download the complete paper here


The Differences Between the North American Building Constructions and the Southern European Building Constructions

The Differences Between the North American Building Constructions and the Southern European Building Constructions Affects the Type of Fire Pattern That Are Most Commonly Found in a Fire Scene

Giovanni Cocchi, P.E., Ph. D.
ARSON Fire, Safety, and Environmental Investigations S.r.l., Italy

Presented at the International Symposium on Fire Investigation Science and Technology, 2014

ABSTRACT

Fire pattern analysis is a fundamental step in any fire investigation, since it provide the basis for the heat and flame vector analysis onto which the reconstruction of fire spread and the identification of the area of origin of the fire are based. Fire patterns formation depends on the fire effects being produced by the fire. This works discuss some of the differences between building construction in North America and Southern Europe, on the basis of the example of Italy, and try to tackle the problem if such differences may affects the type of fire patterns that are found in a fire scene.

Download the complete paper here


Arc Mapping: A Review of Findings and a Reply to the ATF Laboratory

There has been a lot of debate in fire investigation industry recently regarding Arc Mapping.  NAFI’s mission is to increase the knowledge and improve the skills of persons engaged in the investigation and analysis of fires, explosions and arsons, or the litigation that ensues from such investigations. The opinions expressed in this paper do not necessarily reflect the opinion and beliefs of NAFI. Over the next three weeks, we will be sharing different views on Arc Mapping with our members and the industry – it is up to you to draw your own conclusions.

Week 1 – Arc Mapping: New Science or New Myth?
Week 2 – Arc Mapping as a Tool for Fire Investigations
Week 3 – Arc Mapping: A Review of Findings

Abstract
The ATF Laboratory recently issued Special Bulletin No. 2 which offered a series of objections to the arc mapping review presented by this author earlier this year. This paper reviews the original research findings and explains the context of these findings. It then addresses in detail the comments made by ATF, and further addresses the concerns about arc mapping as currently used in the field. It reaffirms that the original findings are correct. It also shows that a new “ellipse” method proposed by the ATF for origin determination from arcing locations is not technically sound.

Download the complete paper here

 


Analysis of Post-Fire Characteristics of Portable Oil Filled Room Heaters

Analysis of Post-Fire Characteristics of Portable Oil Filled Room Heaters to Determine Pre-Fire Orientation

K. Scott Barnhill, PE
Investigative Forensic Specialists, PLLC

Presented at the International Symposium on Fire Investigation Science and Technology, 2014

ABSTRACT
A portable oil filled room heater is a steel vessel typically with 5 to 8 fins containing approximately 3 to 4 liters of mineral oil. The mineral oil acts as the heat transfer fluid that is heated by an immersed electric heating element. The outward appearance of an oil filled heater is that of an old steam radiator.

An oil filled heater (OFH) exposed to full room involvement conditions will commonly display fire effects to include: rupture of the vessel’s spot welds, localized wavy deformations to the fins in the unwetted regions of the vessel, and differential deformation of the vessel’s fins. This study addresses the post-fire characteristics of oil filled heaters exposed to full room involvement conditions in two burn cells. A primary focus of the study is to determine if the pre-fire orientation of the oil filled heater can be determined by analysis of the post-fire appearance. A description of the dynamics that occur between the wetted and unwetted surfaces when the vessel is exposed to full room involvement conditions is discussed.

Analysis of an OFHs’ post-fire characteristics, to include vessel dimensions and wavy deformations, compared to the burn cell experiment results, allows an investigator to accurately interpret the normal reaction of an OFH to full room involvement as well as to determine its pre-fire orientation (upright or otherwise).

Download the complete paper here


Arc Mapping as a Tool for Fire Investigations

There has been a lot of debate in fire investigation industry recently regarding Arc Mapping.  NAFI’s mission is to increase the knowledge and improve the skills of persons engaged in the investigation and analysis of fires, explosions and arsons, or the litigation that ensues from such investigations. The opinions expressed in this paper do not necessarily reflect the opinion and beliefs of NAFI. Over the next three weeks, we will be sharing different views on Arc Mapping with our members and the industry – it is up to you to draw your own conclusions.

Week 1 – Arc Mapping: New Science or New Myth?
Week 2 – Arc Mapping as a Tool for Fire Investigations

ATF Fire Research Laboratory Technical Bulletin
Arc Mapping as a Tool for Fire Investigations

ABSTRACT
The purpose of this Technical Bulletin is to address statements that are contained in Arc Mapping: New Science or New Myth?, which was presented at the Fire and Materials Conference in February 2017. The paper’s interpretations and analysis of the published literature, the limitations it places on the process of arc mapping, and the conclusion that arc mapping is applicable to less than 1% of fire scene investigations are misleading. This bulletin addresses some of these issues.

Download the entire technical bulletin here.

 


Arc Mapping: New Science or New Myth?

There has been a lot of debate in fire investigation industry recently regarding Arc Mapping.  NAFI’s mission is to increase the knowledge and improve the skills of persons engaged in the investigation and analysis of fires, explosions and arsons, or the litigation that ensues from such investigations. The opinions expressed in this paper do not necessarily reflect the opinion and beliefs of NAFI. Over the next three weeks, we will be sharing different views on Arc Mapping with our members and the industry – it is up to you to draw your own conclusions.

Week 1 – Arc Mapping: New Science or New Myth?
Week 2 – Arc Mapping as a Tool for Fire Investigations

Vytenis Babrauskas
Fire Science and Technology Inc., San Diego, CA
Dept. of Mechanical and Aerospace Engineering, University of California San Diego, La Jolla, CA

ABSTRACT
Arc mapping was first introduced in the 2001 edition of NFPA 921 and was subsequently expanded so that in the recent editions it constitutes one of the four main methods for determining the origin of a fire. Careful consideration of engineering principles and large-scale experimental studies on the subject indicates that the relevance and prominence of arc mapping as a leading indicator of fire origin is greatly overstated. The technique is valid and applicable only in some very limited scenarios. Yet it has seen very extensive use in recent years by investigators preparing fire reports. In many cases, such attempted use of arc mapping is based on incorrect and invalid hypotheses, which are often implicitly assumed to be true instead of being explicitly stated. The following are myths: (i) An abundance of arc beads at a given locale means that fire originated in that area, while a paucity of arc beads indicates that it did not. (ii) When multiple arcs are present on a circuit, the direction of arcing will necessarily proceed upstream towards the power source. (iii) If an appliance is the victim of a fire, internal arcing will be primarily near the exterior of the unit, while arcing deep inside indicates a fire origin at that place. NFPA is urged to revise NFPA 921 to eliminate arc mapping as one of the four main methods for establishing fire origin, and to subsume it under the more general category of “fire patterns.” In addition, it is important that NFPA 921 reduce the implied general utility of the method and provide more explicit information on its interpretation and its limitations and on the circumstances under which it may be a valid method for assisting in the determination of the fire origin.

Download the complete paper here


St. John’s School Fire – October 28, 1915

The St. John’s School fire was a deadly fire that occurred on the morning of October 28, 1915, at the St. John’s School on Chestnut Street in the downtown area of Peabody, Massachusetts. Twenty-one girls between the ages of 7 and 17 were burned or crushed to death while attempting to escape the fire.

More than 600 children were in the building when the fire began in the basement of the school building. There were no fire escapes on the outside of the building, but instead those inside were forced to use wide stairways at either end of the interior which led down to the front exit. Mother Superior Aldegon, who led the Sisters who taught in the Catholic school, sounded a fire alarm and began the routine fire drill procedure.

This procedure should have led to the children and teachers leaving the building through the stairways to and out of a rear exit. However, as smoke thickened and the fire came closer, they ran for the front door instead, and became jammed in the vestibule. The fire broke through to the vestibule from directly under the front entrance and the vestibule, now crowded with pupils, was enveloped in flames. The fire rapidly swept through the three-story brick and wooden building, fully engulfing it in less than five minutes.

With their exit blocked, many of the children escaped through first-floor windows or jumped from those on the second and third floors. Not all were able to escape, however; the bodies of the 21 victims were found after the fire subsided, huddled together and burned beyond recognition, on the inside of the school entrance. The Sisters of Notre Dame who taught at St. John’s aided the children trying to escape, some by dropping the students into coats and blankets being used as life nets. These actions were credited in saving many lives. Two of the nuns were injured, one suffering serious burns; however, none of the adults were killed.

As a result of this fire, Peabody became the first city to pass a law that said all doors (in public buildings and school) must push out.

Source: Wikipedia


Reduced Scale Enclosure Testing with Low Heat Release Initial Fuel Packages

Mark A. Campbell, CFPS, SET
Wheat Ridge Fire Protection District, USA

Presented at International Symposium on Fire Investigation, 2014

ABSTRACT

The scientific method is the process in which the fire investigator, among other steps, develops a hypothesis and tests it. The 2011 NFPA 921, §20.5.1, states that “Fire testing is a tool that can provide data that compliment data collected at the fire scene (see 4.3.3), or can be used to test hypotheses (see 4.3.6). Such fire testing can range in scope from bench scale testing to full-scale recreation of the entire event.” (bold and italics added). A Full Scale Enclosure (FSE) testing of a hypothesis may be quite expensive, time consuming, and just not practical. Building and burning a Reduced Scale Enclosure (RSE) may provide insight into the various fire effects, patterns, and dynamics within the enclosure.

Previous FSE burns at Eastern Kentucky University examined the results of a low heat release rate initial fuel source and how the area of origin, based upon the fire effects and fire patterns, preserved through post flashover. This paper will discuss current research on the same concept but with the RSE (1/4 scale). Through the use of the applicable scaling laws, low heat release rate initial fuels were designed and applied to various locations on and around the furniture. The RSEs were burned two minutes post flashover. In all four test burns the areas of origin were determined based upon the collective fire effects and fire patterns. These results have demonstrated that the RSE, when applying the scaling laws appropriately, are a very useful tool for fire investigator and fire protection engineers.

Download the complete paper here.


Distinguishing Between Arcing and Melting Damage

DISTINGUISHING BETWEEN ARCING AND MELTING DAMAGE IN ELECTRICAL RECEPTACLES

Matthew Benfer and Daniel Gottuk
Hughes Associates Inc., USA

Presented at International Symposium on Fire Investigation, 2014

ABSTRACT

The majority of fire-investigation related literature on electrical arcing focuses on copper wiring, both stranded and solid, with some attention paid to steel (i.e., conduit), and relatively little mention of brass. This is despite the relatively equal presence of copper, steel, and brass in receptacles and similar electrical devices. Changes to NFPA 921 in the 2014 edition of the guide expand upon the characteristic traits which can be used to assess whether arcing or melting is present in a conductor. However, most of the characteristic traits of arcing and melting are qualitative and not well defined in NFPA 921, which leads to more subjective evaluations. In addition, a myopic examination of evidence with respect to the presence of one or two characteristic traits can lead to a false indication of arcing. In cases such as this, other evidence of melting (i.e., in close proximity to the area in question) could preclude confirmation of arcing.

The purpose of this work was to determine which characteristic traits are effective in assessing potential arcing damage on receptacle components and wiring. A total of 86 receptacles were evaluated in this study. Thirty-nine receptacles failed as a result of an overheating connection resulting in arcing damage; this included 95 individual conductors. Forty-seven receptacles with fire-induced arcing were also evaluated; this included 87 individual conductors. All of the evaluated receptacles with fire-induced arcing were energized or energized with a load during testing. In contrast, thirty-seven non-energized receptacles with fire induced melting were evaluated with 57 individual conductors.

The characteristic arcing traits which were evaluated include: corresponding damage on the opposing conductor; localized point of contact with a sharp line of demarcation between undamaged and damaged areas; round, smooth shape; resolidification waves; tooling marks visible outside the area of damage; internal porosity; spatter deposits; and small beads and divots over a limited area. The characteristic traits of melting which were evaluated include: visible effects of gravity; gradual necking of the conductor; and pitting, thinning, and presence of holes in the conductor. These traits were taken from the literature (e.g., NFPA 921) and from observations made during the forensic examinations of receptacles and wiring conducted as part of this work. For each characteristic, there were three possible outcomes: Yes, No, and Possible. Yes indicated that the characteristic was judged to be present on the particular conductor; no indicated that the characteristic was judged not to be present on the conductor. Possible indicated that confirmation could not be made either for or against the presence of the characteristic. All of the evaluations were conducted by the same person.

Corresponding damage on the opposing conductor, localized damage with a sharp line of demarcation, and tooling marks outside of the area of damage were observed on significant portions of arc damaged conductors and small numbers of conductors with melting damage; these characteristics were found to be strong indicators of arcing.

Using multiple characteristic traits and contextual information for determination of arcing vs. fire-melting provides greater confidence in the evaluation of damage. In addition, visual examinations were found to be reliable indicators of both arcing and fire-melting for most conductors. However, there are some cases which would benefit from more advanced examination techniques including SEM/EDS examinations, X-ray, CT scanning (X-ray computed tomography), cross-sectioning and polishing, or other metallurgical methods.

Download the complete paper here.


Anatomy of a Wrongful Arson Conviction

ANATOMY OF A WRONGFUL ARSON CONVICTION: SENTINEL EVENT ANALYSIS IN FIRE INVESTIGATION

Paul Bieber, CFEI, B.S., M.L.S. The Arson Research Project

Presented at International Symposium on Fire Investigation, 2014

ABSTRACT

Anatomy of a Wrongful Arson Conviction will discuss the first comprehensive review of U.S. arson exonerations and the first application of sentinel-event and root-cause analysis to the field of fire investigation. Its purpose is to expose and explain the common factors that contribute to wrongful arson convictions.

Sentinel-event analysis has been embraced by several industries as an objective method of identifying and explaining the root causes of errors that have led to harmful outcomes. By reviewing dozens of arson cases, the Arson Research Project has documented the common errors at the heart of many fire investigations where accidental, natural or undetermined fires have been misidentified as arson.

This paper will also highlight the presence and impact of various forms of cognitive bias in each case study and emphasize the importance of objectivity and independence in the reliable application of the scientific method.

The 27 cases being reviewed include 19 exonerations, 7 cases where charges were dropped or a jury returned a not- guilty verdict, and one case that resulted in an execution. Together they represent over 200 years of combined incarceration and several life sentences. Even in the cases where the defendant was acquitted or the charges were dropped, the financial cost and emotional toll to the wrongfully accused were enormous. It is only through a clear examination and better understanding of these common errors that we may hope to avoid similar errors. This case- study review will attempt to shed some light on the problem in an ongoing effort to improve the practice of fire investigation and avoid future wrongful arson convictions.

Download the complete paper here.