Research Report Digest, Issue 11

August 2013
In NIJ’s Research Report Digest, you will find brief descriptions of studies in various criminal justice disciplines, such as crime and forensic sciences, and evaluations of technologies that are used in the law enforcement and corrections fields.
This issue includes reports based on NIJ-funded research that were added to the NCJRS Abstracts Database from January–March 2013.
Find research reports related to:

Crime


Assessment of the Comprehensive Anti-Gang Initiative: Final Project Report (pdf, 197 pages)
Authors: Edmund F. McGarrell; Nicholas Corsaro; Chris Melde; Natalie Hipple; Jennifer Cobbina; Timothy Bynum; Heather Perez

The U.S. Department of Justice developed the Comprehensive Anti-Gang Initiative (CAGI) to support local communities in their efforts to prevent and control gang crime. Unfortunately, most the cities that carried out the initiative could not provide consistent and reliable measures of gang crime. So most of the impact analyses focused on violent crime. Overall, the CAGI cities experienced a greater decline in violent crime than the comparison cities, but the difference was not statistically significant when controlling for concentrated disadvantage and population density. When enforcement implementation levels were included, the high-enforcement CAGI cities experienced a statistically significant 15-percent decrease in violent crime. The comparison based on a propensity matching approach yielded similar findings. Among the CAGI cities only, higher levels of federal prosecution for gun crime negatively correlated to violent crime.


Developing an Actuarial Risk Assessment to Inform Decisions Made by Adult Protective Services Workers (pdf, 127 pages)
Authors: Kristen Johnson; Kathy Park; Andrea Bogie; Shannon Flasch; Jennifer Cotter

This study modeled an actuarial risk assessment that would classify people reported for adult maltreatment or self-neglect. The researchers also assessed whether adult protective services workers could use the risk assessment to make decisions related to the likelihood of future maltreatment or self-neglect. Initial findings were promising. However, more research is needed to improve the assessment’s classification abilities. Results from the process evaluation show that this work could be helpful to APS workers. This research shows a need for research-based assessments for APS field staff and managers that can improve the decisions made by APS caseworkers.


Developing the Capacity to Understand and Prevent Homicide: An Evaluation of the Milwaukee Homicide Review Commission (pdf, 95 pages)
Authors: Deborah Azrael, Ph.D.; Anthony B. Braga, Ph.D.; Mallory O'Brien, Ph.D.

The goal of the Milwaukee Homicide Review Commission is to support innovative homicide prevention and intervention strategies. The impact evaluation showed that MHRC interventions produced a statistically significant 52-percent decrease in the monthly homicide rate in the treatment districts. Researchers found a 9.2-percent decrease in the monthly homicide rate in the control districts. The interventions were associated with a noteworthy decrease in homicide.

Violence


In, Out, and In Again? A Life Course Understanding of Women's Violent Relationships (pdf, 204 pages)
Author: Kristin Carmela Carbone-López

This research points to two main policy implications. First, it is important to learn whether a woman is leaving a violent relationship or experiences violence across relationships. This will help researchers identify which strategies successfully help women to escape violence. It will also help them develop specific interventions and treatments to help women cope with their violent relationships. Second, given the potentially disastrous effect of childhood victimization, it is important to identify victims of child abuse and address the abuse while the victims are still young and before they enter their own intimate relationships. In these cases, efforts should be made to both prevent child abuse and preemptively address the lifetime effects of such abuse.


Evaluating the Work of Sexual Assault Nurse Examiner (SANE) Programs in the Criminal Justice System: A Toolkit for Practitioners (pdf, 227 pages)
Authors: Rebecca Campbell, Ph.D.; Megan Greeson, M.A.; Nidal Karim, Ph.D.; Jessica Shaw, M.A.; Stephanie Townsend, Ph.D.

The SANE Practitioner Evaluation Toolkit provides practitioners with the information necessary to understand and evaluate a SANE program’s impact on local prosecution rates. It also offers ideas for using this information to improve the program’s positive impact on the reporting, investigation and prosecution of sexual assault cases. It guides the user through six evaluation steps: understanding the evaluation design, identifying evaluation questions, setting up cooperative agreements, sampling cases and collecting data, analyzing the data, and interpreting the results.


Implementation of a Sexual Assault Nurse Examiner (SANE) Practitioner Evaluation Toolkit (pdf, 145 pages)
Authors: Rebecca Campbell, Ph.D.; Stephanie Townsend, Ph.D.; Deborah Bybee, Ph.D.; Jessica L. Shaw, M.A.; Jenifer Markowitz, N.D., R.N.

This project evaluated implementation of the SANE Practitioner Evaluation Toolkit, which was developed as part of a previous National Institute of Justice-funded study on how SANE programs affect criminal justice system case outcomes. Six SANE programs used the toolkit to evaluate whether it increased prosecution rates. None of the sites experienced a statistically significant increase in prosecution rates following implementation of the SANE program. However, when data were aggregated across sites, the evaluation found that cases processed post-SANE implementation were 80 percent more likely to get a higher level of disposition compared to cases processed before implementation, which was a significant increase. These findings suggest the SANE intervention model has improved sexual assault case progression in the criminal justice system. Nevertheless, the vast majority of both pre‐SANE and post-SANE resulted in nonreferral/no charges filed.

Forensic Sciences


Application of Chemometrics and Fast GC-MS Analysis for the Identification of Ignitable Liquids in Fire Debris Samples (pdf, 45 pages)
Authors: Michael Sigman, Ph.D.; Mary Williams, M.S.

The researchers developed a chemometric method of data analysis to help identify gas chromatography/mass spectrometry (GC-MS) patterns associated with ignitable liquid classes, as designated under ASTM E 1618-10. They also set out to classify ignitable liquid residue in the presence of background interferences found in fire debris. The researchers tested the method in three ways. First, they used computationally constructed datasets prepared using total ion spectra from ignitable liquids and pyrolyzed substrates from the Ignitable Liquids Reference Collection and Database and the Substrate Databases. Second, they used laboratory-generated burn samples. Third, they used large-scale burn samples produced specifically for testing purposes.

The first set of tests showed that ignitable liquid contributions can be detected, even as a minor component, in common building materials. Results from the laboratory-scale burns showed an 80- to 85-percent success rate for classification. The large-scale burns conducted in the third task achieved similar results.


Application of Fluorescence Line Narrowing Spectroscopy to Forensic Fiber Examination (pdf, 144 pages)
Authors: Andres D. Campiglia, Ph.D.; Michael D. Sigman, Ph.D.

This project developed an identification tool for textile fibers that preserves the evidentiary value of the original sample by focusing on the total fluorescence emission of fibers. It also examined the contribution of the textile dye (or dyes) to the fluorescence spectrum of the fiber, and the contribution of intrinsic fluorescence impurities as a reproducible source of fiber comparison. The proposed method, instrumentation and data analysis reduce the need for destructive dye extractions while significantly increasing the discriminating power of fluorescence microscopy.


Application of Spatial Statistics to Latent Print Identifications: Towards Improved Forensic Science Methodologies (pdf, 109 pages)
Authors: Stephen J. Taylor; Emma K. Dutton; Patrick R. Aldrich; Bryan E. Dutton

This project reviewed the latent fingerprint ACE-V comparison method to ascertain the fingerprint features considered during the comparison process and to apply principles of spatial analyses to calculate false-match probabilities. The objectives were to spatially analyze fingerprint features using Geographic Information Systems (GIS) techniques and empirically derive probabilities to provide a quantitative measure of the discriminating value of the various ridgeline features.

GIS results showed a greater density of minutiae and ridgelines below the core than above the core, regardless of pattern type. However, the distributions of bifurcations and ridge endings were more similar within any pattern type rather than across different pattern types. Also, pattern types with comparable ridge flow (e.g., right and left slant loops, and whorls and double-loop whorls) had greater similarity between them when comparing various metrics such as axis dimensions and Thiessen polygon ratios. Geometric morphometric (GM) results showed little shape variation among fingerprints of the same pattern type, with the greatest shape variation associated with the deltas. Additional GM spatial analyses suggested a high degree of shape consistency between left and right slant loops and between whorls and double loop whorls. Monte Carlo simulations showed the probability of random minutiae correspondence drastically decreased as the fingerprint attribute criteria (e.g., minutiae type, direction) increased. In addition, increasing the number of minutiae and fingerprint attributes applied in searches away from the core and delta regions yielded lower probabilities for a false match. However, minutiae spatial distributions in regions around and below the core were not always unique.


Automated Processing of FTA Samples (pdf, 27 pages)
Authors: Erin Finehout; Perry Bonanni; Scott Duthie; Weston Griffin; Zaeem Khan; Phil Shoemaker; Xuefeng Wang

Forensic laboratories use fast technology for analysis of nucleic acids (FTA) cards to collect and store DNA samples. This project designed an automated system for processing FTA cards for STR analysis in forensic DNA databanking labs. The system allows a user to load a stack of FTA EasiCollect cards into the system and retrieve a 96-well plate of punches ready for polymerase chain reaction (PCR) analysis. The system reduced the risk of sample loss from static effects, with a final estimated error rate of less than 0.15 percent. Tests to evaluate cross-contamination show that a cleaning punch between samples reduces the cross-contamination risk, and the current vacuum system needs improvement to prevent general dust buildup. In a final test, researchers punched FTA cards containing buccal swabs. All the samples tested yielded full short tandem repeat (STR) profiles.


Automated Processing of Sexual Assault Cases Using Selective Degradation (pdf, 75pages)
Authors: Dr. Christian Carson; Dr. Alex Garvin; Kim Gorman

This research addressed the problems of getting a single-source sperm DNA profile from sexual assault evidence that contains both sperm cell and epithelial cell DNA and of automating the differential extraction process. The resulting process for getting single-source male DNA profiles from mixed stains improves the ability to process and prosecute sexual assault cases. By eliminating sperm fraction DNA profile mixtures from almost all sexual assault evidence containing sperm cells, crime laboratories will often be able to get single-source male DNA profiles when the standard method of differential extraction does not produce a sperm fraction male profile. This process saves time in performing differential extractions and removes the uncertainty caused by mixed DNA profiles. Also, by automating the differential extraction process, it will help reduce the backlog of sexual assault evidence.


Cognitive and Contextual Influences in Determination of Latent Fingerprint Suitability for Identification Judgments (pdf, 36 pages)
Authors: Peter Fraser-Mackenzie; Itiel Dror; Kasey Wertheim

This study examined forensic fingerprint examiners' suitability determinations of latent fingerprints. It compared situations in which the latent print was assessed by itself to situations in which it was presented alongside a comparison matching or nonmatching exemplar print. Examiners were more inclined to conclude that the latent print was suitable for comparison when presented alongside a nonmatching comparison exemplar print than when it was presented alone, even when the latent print was unsuitable for comparison. When presented alongside a matching comparison exemplar, examiners were more likely to classify the latent print as questionable, compared to solo analysis. This effect persisted even when the latent presented was suitable, suggesting a strong main effect.

Additionally, if one examiner determined the latent print was unsuitable and a second examiner knew this, that examiner would be more likely to come to the same conclusion. However, knowledge that an examiner determined the print to be "suitable" did not increase the likelihood that other examiners would feel the same way. These effects were weaker — although not removed — in examiners with International Association for Identification certification, suggesting that training may reduce the effects of influence and bias in suitability determinations. In addition, latent prints that had been classified as "unsuitable" in a non-biasing context also tended to be classified as "unsuitable" when presented in a strongly biasing context.


Day and Night Real Time Signature Enhanced Crime Scene Survey Camera (pdf, 78 pages)
Authors: Milind Mahajan; Karen Zachery; Weiya Zhang; Andrew Brackley; Xindian Long

This project developed an integrated day and night forensic survey camera that can improve targets of forensic interest in real time. The camera collects and analyzes different wavelength and polarization channels using multiple wavelength illumination, including ultraviolet for fluorescence. Study findings showed contrast improvement and improved detection threshold across a wide range of targets. In several cases, the camera captured target imagery (e.g., prints, stains) that was invisible to the naked eye and to conventional color digital cameras. Adaptive dynamic range compression and false color rendering were effective in directing viewers’ attention to the improved target.


Denver DNA Efficiency Improvement Project, Final Technical Report (pdf, 97 pages)
Authors: Lindsey R. Horvat; Susan G. Berdine; Greggory S. LaBerge, Ph.D.

This project created a simulated model of the workflow of the Denver Forensic Biology/DNA Unit using specialized software and identified areas for efficiency improvements. It also drew on employee input and a teamwork approach to identify and implement additional efficiency improvements that would help decrease the backlog and turnaround time of DNA cases.

The results of test simulations of the model using a year of casework data found that instruments and equipment were not causing bottlenecks in the workflow, and increasing the number of instruments did not improve the case backlog. Rather, increasing the number of trained DNA personnel improved the backlog and turnaround times to the needed levels. Also, the simulation provided definitive data on the number of staff needed to eliminate the case backlog and meet target objective turnaround times.


Developing a High Throughput Protocol for Using Soil Molecular Biology as Trace Evidence (pdf, 90 pages)
Authors: Sabreena Larson; David O. Carter; Cheryl Bailey; Rhae A. Drijber

This research was designed to find out the optimal conditions for storing soil when using biochemical molecules as trace evidence. First, researchers analyzed DNA profiles made with capillary electrophoresis-single-stranded conformation polymorphism (CE-SSCP). They found that storage treatment did not have a significant impact on the genetic profile of soil bacteria. However, there were significant differences between soils collected from different sites. This shows that different soils can have different genetic profiles and that soil storage has little effect in their characterization.

Next, fatty acid profiles were analyzed as fatty acid methyl esters. Researchers found -80˚C to be the best storage temperature to preserve the integrity of the microbial community fatty acid profile, although -20˚C was acceptable. The researchers infer that analysis of fatty acid profiles is more sensitive to changes in the microbial habitat than that of DNA profiles. In contrast, genetic changes in the microbial community occur over longer periods.


Development and Validation of a Method for Individualization of Middle Petroleum Distillates and Kerosene Ignitable Liquids (pdf, 109 pages)
Authors: J. Graham Rankin, Ph.D; Peter Harrington, Ph.D.

This project created analytical tools that provide a statistical evaluation for forensic comparison between ignitable liquid residues in fire debris and ignitable liquids that a suspect has. The project applied target compound analysis from high-resolution Gas Chromatography/Mass Spectrometry (GCMS), used for gasoline individualization, to medium petroleum distillates (MPD) and kerosene, which may be used as accelerants in arson cases. For kerosene, the authors applied the projected difference resolution (PDR) mapping technique to measure the quantitative differences among the ignitable liquid samples by their GCMS profiles, and applied fuzzy rule-building expert systems (FuRES) to classify individual ignitable liquids. The FuRES models yielded correct classification rates greater than 90 percent for discriminating between samples. PDR mapping, a new method for characterizing complex datasets, was consistent with the FuRES classification result. For MPD, the authors used Kendall’s-tau metric for “association/no-association” to show “association/no-association” true-positive rates more than 95 percent, with false-positive rates of about 5 percent.


Development of Synthetically Generated LEA Signatures to Generalize Probability of False Positive Identification Estimates (pdf, 72 pages)
Authors: Benjamin Bachrach; Pan Gao; Roger Xu; Wei Wang; Ajay Mishra; Kaizhi Tang; Guangfan Zhang

Automated ballistics identification systems use a land-engraved area (LEA) signature, a one-dimensional signal computed from a digitized three-dimensional bullet surface, to match bullets. This study developed a bullet data-synthesis approach to producing LEA signatures by capturing the characteristics of each firearm brand and the details of each barrel. Preliminary results show that newly generated LEAs for a brand are similar to other LEAs of the same brand but dissimilar to those of different brands. Also, the correlation distributions for matched and nonmatched LEAs of the same brand suggest the combined data approaches the same distribution as the experimental data.


DNA Profiling of the Semen Donor in Extended Interval Post-Coital Samples (pdf, 61 pages)
Author: Jack Ballantyne, Ph.D.

Despite improved extraction and profiling techniques, forensic examiners still routinely fail to recover male-donor DNA profiles from samples collected from living sexual assault victims more than 6 days after intercourse. This study developed new strategies to enable the recovery of this DNA from extended interval post-coital samples. Using a combination of novel methods to improve male DNA fractions, researchers were able to get male donor profiles in samples collected 6 to 9 days following intercourse. This significantly increases the timeframe in which male profiles can be successfully recovered from post-coital samples.


Elemental Analysis of Glass by SEM-EDS, uXRF, LIBS and LA-ICP-MS (pdf, 88 pages)
Authors: Jose Almirall; Ben Naes; Erica Cahoon; Tatiana Trejos

This project sought to improve the value of trace evidence examinations, including mature analytical techniques such as laser-ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) for glass and paint. Additionally, researchers sought to develop a database of elemental data for more than 700 different glass samples that shows the significance of a trace elemental “match.” Forensic examiners can use this information to interpret comparisons of materials using trace elemental composition more judiciously.

The project also evaluated the performance of laser-induced breakdown spectroscopy (LIBS) in analyzing glass. Researchers developed an analytical scheme for using LIBS, which they then validated and used to analyze many glass samples. LIBS provided excellent discrimination potential for a glass set of 41 automotive fragments recovered from 14 different vehicles. Researchers compared the discrimination power of LIBS to two of the leading elemental analysis techniques, uXRF and LA-ICP-MS, with similar results. All three methods produced greater than 99-percent discrimination, and the pairs found indistinguishable were correlated across the analytical methods.


Estimation of Age at Death Using Cortical Bone Histomorphometry (pdf, 87 pages)

Author: Christian Crowder, Ph.D.

Biological variability in age indictors and different skeletal responses to environmental factors over an individual’s life make it difficult to estimate the age at death of the adult skeleton, particularly for people older than age 50. Previously developed methods using the femur showed significant methodological issues that affect their reliability and accuracy. This research evaluated histological age estimation using the anterior femur and explored the biological limitations of bone turnover as an age indicator. The results indicate that histological analysis of the anterior femur provides reliable age estimates, and that the described regression model is most accurate for people over age 50.


Estimation of Biological Parameters for Human Identification in Cases of Missing Persons, Mass Disasters, and Human Rights Violations (pdf, 354 pages)
Authors: Erin H. Kimmerle, Ph.D.; John Obafunwa, M.D.; J.D. George Kamenov, Ph.D.; Lyle W. Konigsberg, Ph.D.

This research project had four primary objectives. The first was to collect osteometric, morphological, elemental and isotope data on skeletal and dental traits for contemporary African populations in Nigeria and other contemporary American populations and Diaspora. The researchers then sought to assess the applicability of methods used across populations by testing the degree of population variation among diverse populations. Next, they recalibrated biological parameters for identification methodology using a Bayesian statistical approach and hazard analysis. The final objective was to develop internationally relevant data protocols and a framework for research in human identification.

The research produced unique morphological and isotopic parameters for human identification. It provides insight into human skeletal variation, with implications for human identification. The findings also show the craniometrical variation among broad African regions and Diaspora populations, and provide the population-specific parameters needed for sex estimation for diverse African groups.


Evaluation and Application of Polynomial Texture Mapping (PTM) in the area of Shoe/Tire Impression Evidence (pdf, 33 pages)
Authors: James S. Hamiel; John S. Yoshida

Researchers built a laboratory-based polynomial texture mapping(PTM) dome, bought and tested camera equipment and developed software to synchronize the digital camera control (image capture) with the flash sequence. Then, they tested the PTM technique with several different types of impression evidence.

When compared to traditional photography techniques, the PTM technique improved the visualization of unique characteristics in several of the impressions captured during testing. Although many of these characteristics were visible in traditional photographs, the detail was not as well defined. The PTM technique allowed the examiner to confirm the detail in the impression corresponded to a unique characteristic in the shoe. However, the use of the PTM technique did not improve the visualization of texture in all the impressions captured when compared to traditional photography techniques.

Forensic Pathology Tool to Predict Pediatric Skull Fracture Patterns (pdf, 130 pages)
Authors: Brian J. Powell; Timothy G. Baumer; Nicholas V. Passalacqua; Christina D. Wagner; Roger C. Haut; Todd W. Fenton; King H. Yang

The researchers conducted impact experiments on developing porcine skulls to show characteristic fracture patterns. They did tests to discover the elastic and failure properties of the developing bone and suture. The researchers also developed a computational tool to predict fracture patterns, and they developed a mathematical model of the pediatric human skull. Research on the porcine animal model showed that with increased impact energy, there were more sites of fracture initiation and propagation into adjacent bones. Fracture initiation occurred away from the impact site. One impact often produced two or more linear fractures away from the impact site. Rigid and compliant interfaces produced different fracture patterns. Additionally, developmental changes in the material properties of porcine skulls from two to 24 days paralleled those of the human skull from two to 24 months. Equivalent energies of impact to an entrapped head produced more fractures than a free-falling head onto a rigid interface.

Some studies on finite-element modeling show that a simplified model of fracture prediction based on principal stress and strain theory can predict general patterns of skull fracture. However, researchers found that more sophisticated models of fracture initiation and propagation were too sensitive to subject-specific characteristics and model assumptions to limit the utility of current computational tools to predict fracture patterns for general use in forensic cases.

Goodness-Of-Fit Tests and Function Estimators for Receiver Operating Characteristic (ROC) Curves: Inference From Perpendicular Distances (pdf, 284 pages)
Author: R. Bradley Patterson

Statistical inference with the receiver operating characteristic (ROC) curve — a tool to assess performance of classification methods — has traditionally been based on differences between empirical and parametric curves in the vertical direction. Working along directions perpendicular to parametric binormal ROC curves, researchers designed a goodness-of-fit test similar to existing statistics based on the empirical distribution function for a single random variable. After first showing the poor performance of the direct application of the Cramér-von Mises family of goodness-of-fit statistics to ROC curves, the new test showed uniformity of p-values under the null hypothesis, proving its consistency. More simulations with alternatives show the goodness-of-fit test’s power in three general cases. First, the test can reject the binormal model estimated with a fully parametric approach and small samples. Second, the test can reject the binormal model estimated with either a semiparametric or fully parametric approach and large samples. Third, the test can reject the binormal model with small samples if one of the underlying distributions follows a truncated, mixture or heavy-tailed distribution.

Improved Detection of Male DNA in Post-Coital Samples (pdf, 106 pages)
Authors:Jack Ballantyne, Ph.D.; Angela van Daal, Ph.D.; Helge Lubenow, Ph.D

Samples collected from sexual assault or homicide victims can contain low levels of cellular male DNA mixed with many female epithelial cells. This often results in failure to get an autosomal short tandem repeat (STR) profile from the male DNA donor. Y-STR analysis can overcome this problem; however, there are still many instances where that approach does not work. The authors developed a 17-locus Y chromosome-specific nested polymerase chain reaction (PCR) preamplification multiplex and performed initial validation studies to show its potential use with forensic samples. Using the nested PCR preamplification prior to Y-STR analysis allows Y-STR profiles to be recovered from as little as 5pg of male DNA (~ 1 diploid cell) from blood, semen, saliva and skin. The presence of female DNA caused no interference, even when it was more than 100,000 times the amount of male DNA. The ability to recover Y-STR profiles from touch/contact DNA samples as well as from extended interval post-coital samples (i.e., more than 5 days after intercourse) shows the potential efficacy of the nested PCR preamplification multiplex for casework.

Improving Sex Estimation From Crania Using 3-dimensional CT Scans (pdf, 80 pages)
Authors: Richard Jantz; Mohamed Mahfouz; Natalie R. Shirley; Emam Abdel Fatah

Estimations of the sex of humans based on their craniums almost exclusively rely on measurements or observations of external cranial features and seldom exceed 90-percent accuracy. This project examined cranial measurements to obtain the best discriminators for sex estimation. Results show that although size is a significant component of cranial sex dimorphism, shape plays a role as well. The global analysis showed significant size differences in cranial length and facial breadth and significant shape differences in the glabellar, zygomatic, occipital and mastoid regions. The local analysis corroborated the global analysis. Important size-related variables were maximum cranial length, cranial base length and mastoid height. The shape-related variables capture differences in the projection of the glabellar region, inclination of the frontal bone, and angulation/orientation of the mastoid. In addition, vault thickness is a sexually dimorphic feature, with females having, on average, thicker vaults than males in the frontal region, and males having thicker vaults in the occipital region (this variable offered high discriminatory power).

The best model is an 11-variable model, which achieved 97.3 percent accuracy. The eight-variable and three-variable models were also successful, with accuracy levels of 95.5 percent and 86.5 percent. The glabella projection index alone classified with 82.4-percent accuracy, whereas the bizygomatic breadth achieved 83 percent accuracy, and the basion-nasion length achieved 82 percent accuracy.

NCFS Support of SWGDE and T/SWGFEX Projects/Activities - T/SWGFEX Portion (pdf, 91 pages)
Authors: Michael Sigman, Ph.D.; Mary Williams, M.S.; Carrie Whitcomb

This project sought to develop or improve three databases used extensively in fire and explosion analysis. The researchers first improved the ignitable liquids reference collection and database and the associated substrates database casework samples. The researchers then developed a smokeless powders database and a computer-searchable ignitable liquids database.

Physical Matching Verification (pdf, 61 pages)
Authors: Yoram Yekutieli; Yaron Shor; Sarena Wiesner; Tsadok Tsach

This project developed a prototype system for physical matching in two dimensions to help forensic experts perform physical matching objectively and to collect statistics and build confidence levels about physical matches. The statistical results were much lower than first expected because only the two-dimensional fracture lines were used, with none of the added clues commonly used in tool mark comparisons. The pairs were also classified as matches and nonmatches; “inconclusive” pairs were ignored. For these reasons, the results are the entry point to numerical or quantitative evaluation. Adding the surface details or performing a three-dimensional match can bring the results much closer to those achieved by toolmark experts.

Quantitative Measures in Support of Latent Print Comparison (pdf, 63 pages)
Author: Sargur N. Srihari

Latent prints of friction ridge impressions have long been useful in identification. The method of examining latent prints — known as ACE-V (analysis, comparison, evaluation and verification) — has been well documented. Several recent influential reports have explained the need to quantify confidences within ACE-V to strengthen the science of friction ridge analysis. This research evaluated three quantitative measures — rarity of features, confidence of opinion and a probabilistic measure of similarity. The rarity of observed features is difficult to compute due to the large number of variables and high data requirements. The proposed solution uses probabilistic graphical models to represent spatial distributions of fingerprints represented at level 2 detail (minutiae). Determining the confidence of opinion is relevant to the evaluation phase of ACE-V. The proposed computation determines a likelihood ratio as a product of rarity and the probability of similarity under the identification hypothesis. The developed methods for measuring similarity probabilistically, and in a manner analogous to cognition, can be useful in examiner training, presentation of opinion and validating examination procedures.

Law Enforcement

New Approaches to Understanding and Regulating Primary and Secondary Illegal Firearms (pdf, 177 pages)
Authors: Glenn Pierce; Anthony Braga; Garen Wintemute; Matthew Dolliver

This report confirms previous study results showing the usefulness of trace data in policing illicit firearms. The data show that jurisdictions with gun regulations recover fewer illicitly distributed weapons compared to jurisdictions without regulations. The authors suggest that more regulation will reduce the availability and distribution of illicit firearms.

Strategies for Disrupting Illegal Firearms Markets: A Case Study of Los Angeles (pdf, 91 pages)

Authors: Greg Ridgeway; Glenn L. Pierce; Anthony A. Braga; George Tita; Garen Wintemute; Wendell Roberts

The primary goal of this project was to find out whether a data-driven, problem-solving approach could yield new interventions aimed at disrupting the workings of local, illegal gun markets serving criminals, gang members and juveniles in Los Angeles. The authors created a new software tool to help law enforcement analyze patterns in crime-gun data and identify and trace illicit pathways by which criminals get guns. They incorporated the findings into an interagency working-group process that developed a community-based intervention designed to disrupt the illegal flow of guns to Los Angeles-area criminals. This intervention may reduce the number of straw purchases. The authors then assessed the utility of retail ammunition-purchase records in identifying prohibited firearm possessors, recommending a cost-benefit analysis on this measure.

Through the Wall Standoff Detection and Tracking of Individuals (pdf, 70 pages)
Authors: AKELA

The goal of this project was to develop an easy-to-use, battery-operated, FCC-compliant, portable sense-through-the-wall standoff radar imaging system that weighs less than 15 pounds and costs less than $5,000. The system must be capable of detecting personnel behind an 8-inch thick concrete block wall at 30 meters; controllable by a wireless Ethernet connection, allowing remote deployment and operation; and capable of producing images identifying stationary and moving individuals within building structures. A device has been developed that detected and tracked stationary and moving objects and human subjects within a 10.5-meter thick double reinforced concrete structure. Final Federal Communication Commission certification testing and a user performance evaluation will occur under an NIJ fiscal year 2012 grant.

Courts and Corrections

Advanced Behavior Recognition in Crowded Environments (pdf, 218 pages)
Authors: Ming-Ching Chang; Weina Ge; Nils Krahnstoever; Ting Yu; Ser Nam Lim; Xiaoming Liu

This program is designed to increase situational awareness in law enforcement and correctional settings and reliably detect and prevent disorderly conduct and criminal behavior. The program stresses developing a robust probabilistic event modeling framework that takes the uncertainty of low-level image evidence into consideration. The event-explanation and scenario-modeling graphical user interface also makes the program user-friendly. The program has led to developing intelligent video capabilities that can help law enforcement and corrections detect many different types of events and, often, can alert operators to the onset of an event, enabling early detection and possible prevention of critical events.

Evaluating A Presumptive Drug Testing Technology in Community Corrections Settings (pdf, 58 pages)
Authors: Craig D. Uchida; Gordon A. Aoyagi; W. Riley Waugh; Shawn Flower; Shellie E. Solomon; Jonathan Mash

This evaluation sought to find out whether presumptive drug-detection technology would increase community corrections agencies’ success in identifying offenders or settings that have been exposed to drugs and decrease the overall cost of drug testing. In addition, the researchers evaluated the cost-effectiveness of the technology.


Measuring the Effect of Defense Counsel on Homicide Case Outcomes (pdf, 59 pages)
Authors: James M. Anderson; Paul Heaton

Since 1993, every fifth murder case tried in Philadelphia, Pa., has been sequentially assigned to the Defender Association of Philadelphia. The other four cases are assigned to appointed counsel. This study sought to measure the difference a lawyer makes to the outcome of a serious criminal case. Compared to private appointed counsel, public defenders reduce the murder conviction rate by 19 percent, reduce the likelihood that their clients receive a life sentence by 62 percent, and reduce overall expected time served in prison by 24 percent. This suggests that defense counsel has an enormous effect on case outcomes.

To understand the difference in outcomes, researchers interviewed judges, appointed counsel and public defenders in Philadelphia. The results show that compared to the public defenders, appointed counsel are impeded by limited compensation, incentives created by that compensation, relative isolation, and conflicts of interest among both the appointing judges and the appointed counsel.

National Institute of Justice's Evaluation of Second Chance Act Adult Reentry Courts: Program Characteristics and Preliminary Themes from Year 1 (pdf, 24 pages)
Author: Christine Lindquist; Jennifer Hardison Walters; Michael Rempel; Shannon M. Carey

In fiscal year 2010, the National Institute of Justice funded a cross-site evaluation of eight Bureau of Justice Assistance-funded reentry courts. The evaluation consists of three phases — a process evaluation, an impact evaluation and a cost-effectiveness study. Findings from the Year 1 process evaluation show that most sites share several common programmatic characteristics. They include an emphasis on post-release service delivery, provision of services relevant to the target population (with all sites offering substance abuse treatment and employment services), a case management approach to coordinating and monitoring services, use of court hearings to monitor participants’ progress in the program, testing for drugs, and a team approach to decision-making about sanctions and rewards.

Predicting Erroneous Convictions: A Social Science Approach to Miscarriages of Justice (pdf, 434pages)
Authors: Jon B. Gould; Julia Carrano; Richard Leo; Joseph Young

This study compared cases in which innocent defendants were wrongfully convicted to "near misses" – cases in which an innocent defendant was acquitted or had charges dismissed before trial. Researchers found 10 significant factors that can lead to a wrongful conviction. They include a younger defendant, a defendant with a criminal history, a weak prosecution case, a prosecution that withholds evidence, a dishonest non-eyewitness, unintentional witness misidentification, misinterpreting forensic evidence at trial, a weak defense, a family witness offered by the defendant, and a "punitive" state culture.


Unobtrusive Suicide Warning System, Final Technical Report (pdf, 151 pages)
Authors: Jeffrey M. Ashe; Meena Ganesh; Lijie Yu; Catherine Graichen; Ken Welles; Bill Platt; Joy Chen

Despite many improvements, inmate suicide remains a longstanding problem for correctional institutions. In phases I and II of this program, a prototype demonstration system was developed that can measure an inmate’s heart rate, breathing and general body motions without being attached to the inmate. Building on the success of the first two phases, a final phase is proposed to design a “hardened” system for long-term use.


UWB Enhanced Time Difference of Arrival System, Final Report (pdf, 32 pages)
Authors: Benjamin Lonske ; Eric van Doorn; Satya Ponnaluri; Arvind Bhat

Corrections officials have identified the problem of cell phones in prisons as one of the toughest issues they face. The purpose of the proposed technology development and evaluation effort is to design, implement and test a prototype time-difference-of-arrival system capable of detecting and locating cell phones in correctional facilities in real time. Test results show that signals received outside a facility can be stronger than those received inside, depending on the proximity of the transmitter and receiver. In addition, high-frequency signals attenuate more than low-frequency signals, and when the receiver and transmitter are far apart, the received signal is weak and the spectrum analyzer does not capture the corresponding wideband code-division multiple-access (WCDMA) and personal communications service (PCS) signals. (WCDMA and PCS are two popular cell phone frequency bands.)