Marshall University forensics professor receives federal grant to analyze interpretation of fire debris

Special to HNN Provided by Marshall University
Dr. J. Graham Rankin
Dr. J. Graham Rankin

HUNTINGTON, W.Va. – Marshall University has received a $540,752 grant from the National Institute of Justice for a two-year project to study factors that affect interpretation of data by fire debris analysts and to develop a computer program to aid in interpretation.

Dr. J. Graham Rankin, a professor of forensic chemistry in the Forensic Science Graduate Program at Marshall, is conducting the study, which began January 1, 2011.

A National Academy of Sciences (NAS) report released in 2009 on the practice of forensic science recommended more basic research to determine the reliability of many tests – like fire debris analysis – that depend on pattern recognition. Rankin said the grant program is a positive response to the NAS report.

He said the study will help fire debris analysts in crime laboratories and private laboratories better understand how to interpret their results. Fire debris analysts work closely with fire debris investigators in local fire departments to determine whether a fire was accidental or intentional.

“Our research will aid in improving the understanding of the accuracy and reliability of the data commonly used by fire debris analysts, and we will be validating techniques,” Rankin said. “This interpretation will be used to determine the presence and classification of ignitable liquid residues found in fire debris which may indicate that the fire was deliberately started.”

For the study, ignitable liquids such as gasoline, kerosene, charcoal lighters and other commonly used accelerants will be used to ignite a variety of wood products and carpeting found in homes.  The fire debris generated will be analyzed by two standard methods used by the forensic community.

Data produced by these methods will be distributed to fire debris analysts across the country as “blind case files” for determination about whether or not an ignitable liquid is present and to identify its classification.

Preliminary analyses performed this summer by Amanda Heeren, a second-year graduate student, indicate that the type of wood as well as the extent of charring are important factors in chromatographic patterns from the standard methods. In February, Hereen will present her work at the American Academy of Forensic Sciences national meeting in Chicago. She continues to work on the research project this academic year.

Statistical analysis of the results will be used to determine the number of “false positives” (a conclusion that an ignitable liquid is present, when none is), “false negatives” (concluding that no ignitable liquid is present when one was used) and any incorrect classification of the residue.

“Because background compound products in fire debris are frequently formed which can appear as low levels of ignitable liquids, most lab protocols require a significant amount present to ‘make a call.’ One factor we are investigating is the minimal amount of ignitable liquid residues needed to make a correct assignment to one of the classes of ignitable liquids as specified by the standard method used,” Rankin said. “One other important factor is that the presence of an ignitable liquid does not mean it was used as an accelerant in an intentional fire. It could be incidental, like residual paint thinner in a freshly painted wall, or maybe the cause of an accidental fire, like gasoline fumes ignited by a hot water heater pilot in an enclosed garage.”

A co-principal investigator on the grant is Dr. Nicholas Petraco, associate professor in the John Jay College of Law, City University of New York, in New York City. Petraco and his students will be responsible for the advanced statistical analysis of interpretation of chromatographic data and the development of the expert system. The same ‘blind case files’ will be used to test the expert system.

The activities for the study, “Interpretation of Ignitable Liquid Residues in Fire Debris Analysis:  Effect of Competitive Adsorption, Development of an Expert System and Assessment of the False Positive/Incorrect Assignment Rate,” are funded by project number 2010-DN-BX-K272 through the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. Research conducted by Hereen was supported by project number 2008-DN-BX-K146 through the National Institute of Justice.