Summary of Murdock et. Al

Introduction

The writers keep reiterating in the introduction that the explanations to date in reference to a statistical model (being established) dealing with firearm and toolmark analysis are fallacies and lack scientific methodology / basis. However, it is stated in the reading that firearm and toolmark examiners have been diligent in searching for a statistical basis for analyzing their work but there is no “robust statistical model” that is comprehensible and reliable. The writers further implicate that there are no reliable standards to help identify and explain firearm and toolmark individuality but there are “theoretical models”. There is inference that determining the probability of matching firearm and toolmark evidence data is sketchy at best and lacks predictability.

Review of the Current Literature in Statistical Applications for Firearms and Toolmarks

This section of the reading gives the reader an evolutionary timeline into the development of techniques for firearm and toolmark identification. The first of this begins with the work of Gunther and Gunther collaborating with Hatcher between 1932 - 1935, formulating the idea that firearm and toolmark data would need to be collected and analyzed by competence researchers to help develop RMPs. Further work was conducted in 1949 by O’Hara and Osterburg by assigning numerical values to toolmarks. Another researcher, Churchman, also completed similar work but was criticized by the article’s authors for not providing proof to substantiate his work. Biasotti completed studies in 1959 by recording the numbers of matching striations after he recognized that there was a need for the empirically based statistical studies. Howitt, Tulleners et al (2008) published a paper that proposed developing a model based on correspondence probabilities. Some of the more recent implementation of analytical machinery / techniques involves the use of: confocal microscopy, laser scanners, comparison computer programs using 3D toolmark surface data, correlation calculations, quantitative consecutive matching striae identification criteria (QCMS), and the application of multiple statistical pattern recognition methods.

Current State of the Use of RMPs in Firearm and Toolmark Identification

The current state of the use of RMPs in firearm and toolmark identification is summed by the authors as not having a unified model (with the exception of QCMS) to compare the striations and toolmark impressions like there is for DNA forensic profiling. Further discussion entails recognizing the known and comparing it to the potential error that is used within the judicial system. Toolmark examiners feel this is necessary since error needs to be measured and managed. It is questionable, according to the authors, whether that RMP can be applied to the firearm and toolmark identification that is agreed upon by all. Discussion was given on the feasibility of using RMP for DNA analysis but not firearm and toolmark identification because of the instability of the toolmarks on the samples and that factors that influence them. The expertise of the firearms and toolmarks examiner is recognized as being influential in differentiating and interpreting what is seen on the sample.

Absolute Versus Practical Identification and Subjectivity

The final section of this article deals with the concepts of absolute certainty and subjectivity. Absolute certainty allows no room for interpretation so “practical certainty” is more feasible to the interpretation of the firearm and toolmark data since there are very few things that are “absolute” and the firearm and toolmark community tends to agree. Practical certainty needs to be viewed favorably and not as “an enemy of justice”. Also in consideration is how the subjective aspects of firearm and toolmark should not be seen as being detrimental to the firearm and toolmark examination. Measurements such as bullet weights, caliber sizes, land and groove counts, land width measurements, and barrel twist direction are viewed as being nonsubjective observations. However, it is the judgement of the examiner (which is considered subjective) that determines if there is criteria present for a match / identification.

Title & Abstract Discussion

The title misrepresents the conceptual theme of the paper. The given title indicates that the authors would be discussing the development of ways to apply and analyze firearm and toolmark identification. However, there is general feeling conveyed that the authors are free to offer criticism in the efforts and methods developed to date for firearm and toolmark identification regarding to there is no reliable model to depend upon. It is interesting to note that the authors did not offer to contribute their efforts to help develop a reliable statistical model but chose to use their “subject area expertise” to point out what is wrong with current developmental research to construct a statistical model for firearm and toolmark analysis.

The structure of the abstract does not conform to standard guidelines of what one would
expect to read in an abstract. The information presented is informative however, it does not convey the content of the paper. Some of the questions to be answered in the abstract should have been: (1) what was the purpose / importance of the paper?; (2) What were the results of the study?; and (3) What conclusions were drawn?. The abstract written by Murdock et. al does not address these questions. Even though background information is given, the abstract would seem more appropriate as being part of the introduction section.