law enforcement agencies,
Smallest Geographic Unit:
Date of Collection:
Unit of Observation:
case (Part 1),
individual (Part 2),
evidence record (Part 3)
All homicide cases investigated by the Phoenix Police Department between July 1, 2003 and June 30, 2005 (Part 1). All homicide victims and survivors whose cases were investigated by the Phoenix Police Department between July 1, 2003 and June 30, 2005 (Part 2). All records of evidence collected at homicide scenes by the Phoenix Police Department between July 1, 2003 and June 30, 2005 (Part 3).
administrative records data
Data Collection Notes:
Users should be aware that the datasets can be linked using the IDNUMBER "IDENTIFICATION NUMBER CREATED BY THE PRINCIPAL INVESTIGATOR" variable. Additionally, for Part 2 and Part 3, ICPSR created the RECORDID "UNIQUE CASE IDENTIFIER CREATED BY ICPSR" variable, which is a unique identifier for each line (i.e. record) in that data file. The datasets cannot be linked using the RECORDID variable.
Part 2 contains data on both killed victims and survivors. Users should be aware that Part 2 contains the VICTIMKILLED "KILLED/INJURED VICTIM" variable which identifies whether the victim was killed or survived.
Data from the survey of investigators in the experimental and comparison squads to determine their levels of satisfaction with crime scene specialists and data from the forensic evidence at murder trials component of the study are not available as part of this data collection.
The purpose of the study was to conduct a process and outcome evaluation of the Homicide Clearance Project in the Phoenix, Arizona, Police Department. The primary objective of the Homicide Clearance Project was to improve homicide clearance rates by increasing investigative time through the transfer of four crime scene specialists to the homicide unit.
In 2004, the Phoenix Police Department received a grant from the Bureau of Justice Assistance providing support for the assignment of four crime scene specialists directly to the department's Homicide Unit. Responsibilities of the crime scene specialists were to collect evidence at homicide scenes, prepare scene reports, develop scene diagrams, and other supportive activities. Prior to the project, homicide investigators were responsible for evidence collection, which reduced the time they could devote to investigations.
Transfers of the four crime scene specialists were effective on July 1, 2004. The crime scene specialists were assigned to two of the four investigative squads within the homicide unit. This organizational arrangement provided for a performance evaluation of the squads with crime scene specialists (experimental squads) against the performance of the other squads (comparison squads). The crime scene specialists began on-the-job training immediately upon their transfers by accompanying investigators to scenes and observing the collection and storage of evidence. The months of July and August 2004 were devoted to on-the-job training. By September 2004, the crime scene specialists were able to handle homicide scenes with minimal supervision from investigators, and they had learned how to prepare scene reports documenting the evidence.
During the course of the evaluation, research staff coded information from all homicides that occurred during the 12-month period prior to the transfers (July 1, 2003 - June 30, 2004), referred to as the baseline period, the 2-month training period (July 1, 2004 - August 31, 2004), and a 10-month test period (September 1, 2004 - June 30, 2005). Data were collected on 404 homicide cases (Part 1), 532 homicide victims and survivors (Part 2), and 3,338 records of evidence collected at homicide scenes (Part 3).
The two primary sources of information for the evaluation were investigative reports from the department's records management system, called the Police Automated Computer Entry (PACE) system, and crime laboratory reports from the crime laboratory's Laboratory Information Management System (LIMS). The PACE system was accessed by patrol officers, investigators, crime scene specialists, latent print examiners, forensic scientists, and others with involvement either at the initial scene or during the subsequent investigation for preparation of original and supplemental reports about a homicide investigation. Homicide cases reviewed for the evaluation averaged about 50 supplements per case. LIMS contained detailed information on the results of analysis of forensic evidence conducted by personnel in the crime laboratory.
Part 1 included all homicide cases investigated by the Phoenix Police Department between July 1, 2003 and June 30, 2005, resulting in 404 observations. The baseline period (July 1, 2003 - June 30, 2004) had 195 cases, the training period (July 1, 2004 - August 31, 2004) had 42 cases, and the test period (September 1, 2004 - June 30, 2005) had 167 cases. Part 2 included all homicide victims and survivors whose cases were investigated by the Phoenix Police Department between July 1, 2003 and June 30, 2005, resulting in 532 observations. Part 3 included all records of evidence collected at homicide scenes by the Phoenix Police Department between July 1, 2003 and June 30, 2005, resulting in 3,338 observations.
Mode of Data Collection:
Phoenix Police Department's Police Automated Computer Entry (PACE) system
Phoenix Crime Laboratory's Laboratory Information Management System (LIMS)
Description of Variables:
Part 1, Part 2, and Part 3 each contain variables that measure squad type (experimental or comparison), time period (baseline, training, or test), and whether six general categories of evidence (biological, weapons, trace, fingerprint, impressions, and drug) were collected.
Part 1 contains a total of 18 variables including number of investigators, number of patrol officers at the scene, number of witnesses, number of crime scene specialists at the scene, number of investigators collecting evidence at the scene, total number of evidence collectors, whether the case was open or closed, type of arrest, and whether the case was open or closed by arrest.
Part 2 contains a total of 37 variables including victim characteristics (victim number, age, race, sex, whether the victim was Hispanic, whether the victim received a hospital transport, whether the victim was killed/injured, the manner by which the victim was killed, weapon group, and body location) and motives (argument, retaliation, robbery, gang-related, drug-related, immigration-related, domestic violence, or other motives). Other variables in Part 2 include an instrumental/expressive homicide indicator, whether the case was open or closed, type of arrest, whether the case was open or closed by arrest, number of investigators, number of patrol officers at scene, number of witnesses, and investigative time to closure.
Part 3 contains a total of 46 variables including primary/secondary scene indicator, scene type, number of pieces of evidence, total time at the scene, and number of photos taken. Part 3 also includes variables that measure whether 16 specific types of evidence (fingerprints, hair, gun projectile fragments, drugs, drug paraphernalia, tire tracks, tool marks, other impressions, fibers, cigarette butts, tools, duct tape, paint, clothing, documents, and other evidence) were found and the number of items of evidence that were collected for 13 specific evidence types: latents, DNA swabs, buccal swabs, handguns, other firearms, casings, spent projectiles, live rounds, gunshot residue kits from persons, gunshot residue kits on property, knives, shoeprints, and electrostatic dust print lifter (EDPL) impressions.
Presence of Common Scales:
Extent of Processing: ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of
disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major
statistical software formats as well as standard codebooks to accompany the data. In addition to
these procedures, ICPSR performed the following processing steps for this data collection:
Created variable labels and/or value labels.
Standardized missing values.
Checked for undocumented or out-of-range codes.