Search results

This study used crime count data from the Pittsburgh, Pennsylvania, Bureau of Police offense reports and 911 computer-aided dispatch (CAD) calls to determine the best univariate forecast method for crime and to evaluate the value of leading indicator crime forecast models.

The researchers used the rolling-horizon experimental design, a design that maximizes the number of forecasts for a given time series at different times and under different conditions. Under this design, several forecast models are used to make alternative forecasts in parallel. For each forecast model included in an experiment, the researchers estimated models on training data, forecasted one month ahead to new data not previously seen by the model, and calculated and saved the forecast error. Then they added the observed value of the previously forecasted data point to the next month's training data, dropped the oldest historical data point, and forecasted the following month's data point. This process continued over a number of months.

A total of 15 statistical datasets and 3 geographic information systems (GIS) shapefiles resulted from this study.

The statistical datasets consist of

• Univariate Forecast Data by Police Precinct (Dataset 1) with 3,240 cases
• Output Data from the Univariate Forecasting Program: Sectors and Forecast Errors (Dataset 2) with 17,892 cases
• Multivariate, Leading Indicator Forecast Data by Grid Cell (Dataset 3) with 5,940 cases
• Output Data from the 911 Drug Calls Forecast Program (Dataset 4) with 5,112 cases
• Output Data from the Part One Property Crimes Forecast Program (Dataset 5) with 5,112 cases
• Output Data from the Part One Violent Crimes Forecast Program (Dataset 6) with 5,112 cases
• Input Data for the Regression Forecast Program for 911 Drug Calls (Dataset 7) with 10,011 cases
• Input Data for the Regression Forecast Program for Part One Property Crimes (Dataset 8) with 10,011 cases
• Input Data for the Regression Forecast Program for Part One Violent Crimes (Dataset 9) with 10,011 cases
• Output Data from Regression Forecast Program for 911 Drug Calls: Estimated Coefficients for Leading Indicator Models (Dataset 10) with 36 cases
• Output Data from Regression Forecast Program for Part One Property Crimes: Estimated Coefficients for Leading Indicator Models (Dataset 11) with 36 cases
• Output Data from Regression Forecast Program for Part One Violent Crimes: Estimated Coefficients for Leading Indicator Models (Dataset 12) with 36 cases
• Output Data from Regression Forecast Program for 911 Drug Calls: Forecast Errors (Dataset 13) with 4,936 cases
• Output Data from Regression Forecast Program for Part One Property Crimes: Forecast Errors (Dataset 14) with 4,936 cases
• Output Data from Regression Forecast Program for Part One Violent Crimes: Forecast Errors (Dataset 15) with 4,936 cases.
• The GIS Shapefiles (Dataset 16) are provided with the study in a single zip file: Included are polygon data for the 4,000 foot, square, uniform grid system used for much of the Pittsburgh crime data (grid400); polygon data for the 6 police precincts, alternatively called districts or zones, of Pittsburgh(policedist); and polygon data for the 3 major rivers in Pittsburgh the Allegheny, Monongahela, and Ohio (rivers).

CrimeStat III is a spatial statistics program for the analysis of crime incident locations, developed by Ned Levine and Associates under the direction of Ned Levine, PhD, that was funded by grants from the National Institute of Justice (grants 1997-IJ-CX-0040, 1999-IJ-CX-0044, 2002-IJ-CX-0007, and 2005-IJ-CX-K037). The program is Windows-based and interfaces with most desktop GIS programs. The purpose is to provide supplemental statistical tools to aid law enforcement agencies and criminal justice researchers in their crime mapping efforts. CrimeStat is being used by many police departments around the country as well as by criminal justice and other researchers.

The program inputs incident locations (e.g., robbery locations) in 'dbf', 'shp', ASCII or ODBC-compliant formats using either spherical or projected coordinates. It calculates various spatial statistics and writes graphical objects to ArcGIS, MapInfo, Surfer for Windows, and other GIS packages.

CrimeStat is organized into five sections:

• Primary file - this is a file of incident or point locations with X and Y coordinates. The coordinate system can be either spherical (lat/lon) or projected. Intensity and weight values are allowed. Each incident can have an associated time value.
• Secondary file - this is an associated file of incident or point locations with X and Y coordinates. The coordinate system has to be the same as the primary file. Intensity and weight values are allowed. The secondary file is used for comparison with the primary file in the risk-adjusted nearest neighbor clustering routine and the duel kernel interpolation.
• Reference file - this is a grid file that overlays the study area. Normally, it is a regular grid though irregular ones can be imported. CrimeStat can generate the grid if given the X and Y coordinates for the lower-left and upper-right corners.
• Measurement parameters - This page identifies the type of distance measurement (direct, indirect or network) to be used and specifies parameters for the area of the study region and the length of the street network. CrimeStat III has the ability to utilize a network for linking points. Each segment can be weighted by travel time, travel speed, travel cost or simple distance. This allows the interaction between points to be estimated more realistically.

• Spatial distribution - statistics for describing the spatial distribution of incidents, such as the mean center, center of minimum distance, standard deviational ellipse, the convex hull, or directional mean.
• Spatial autocorrelation - statistics for describing the amount of spatial autocorrelation between zones, including general spatial autocorrelation indices - Moran's I , Geary's C, and the Getis-Ord General G, and correlograms that calculate spatial autocorrelation for different distance separations - the Moran, Geary, Getis-Ord correlograms. Several of these routines can simulate confidence intervals with a Monte Carlo simulation.
• Distance analysis I - statistics for describing properties of distances between incidents including nearest neighbor analysis, linear nearest neighbor analysis, and Ripley's K statistic. There is also a routine that assigns the primary points to the secondary points, either on the basis of nearest neighbor or point-in-polygon, and then sums the results by the secondary point values.
• Distance analysis II - calculates matrices representing the distance between points for the primary file, for the distance between the primary and secondary points, and for the distance between either the primary or secondary file and the grid.
• 'Hot spot' analysis I - routines for conducting 'hot spot' analysis including the mode, the fuzzy mode, hierarchical nearest neighbor clustering, and risk-adjusted nearest neighbor hierarchical clustering. The hierarchical nearest neighbor hot spots can be output as ellipses or convex hulls.
• 'Hot spot' analysis II - more routines for conducting hot spot analysis including the Spatial and Temporal Analysis of Crime (STAC), K-means clustering, Anselin's local Moran, and the Getis-Ord local G statistics. The STAC and K-means hot spots can be output as ellipses or convex hulls. All of these routines can simulate confidence intervals with a Monte Carlo simulation.

• Interpolation I - a single-variable kernel density estimation routine for producing a surface or contour estimate of the density of incidents (e.g., burglaries) and a dual-variable kernel density estimation routine for comparing the density of incidents to the density of an underlying baseline (e.g., burglaries relative to the number of households).
• Interpolation II - a Head Bang routine for smoothing zonal data that can be applied to events (volumes), rates or can be used to create rates. In addition, there is an interpolated Head Bang routine for interpolating the smoothed Head Bang result to grid cells.
• Space-time analysis - a set of tools for analyzing clustering in time and in space. These include the Knox and Mantel indices, which look for the relationship between time and space, and the Correlated Walk Analysis module, which analyzes and predicts the behavior of a serial offender and a spatial-temporal moving average.
• Journey to crime analysis - a simple criminal justice method for estimating the likely location of a serial offender given the distribution of incidents and a model of travel distance. The routine allows the user to estimate a travel model with a calibration file and apply it to the serial events. It can be used to identify a likely location given the distribution of 'points' and assumptions about travel behavior. There is a routine for drawing lines between origins and destinations (crime trips).
• Bayesian journey to crime analysis - an advanced criminal justice method for estimating the likely location of a serial offender given the distribution of incidents, a model of travel distance, and an origin-destination matrix showing the relationship between where crimes were committed and where offenders lived. A diagnostics routine analyzes serial offenders for whom their residence is known and estimates which of several journey to crime estimates is most accurate. A selected method can be applied to identify a likely residence location of a single serial offender given the distribution of incidents, assumptions about travel behavior, and the origin of offenders who committed crimes in the same locations.
• Regression modeling - a module for analyzing a relationship between a dependent variable and one or more independent variables. The CrimeStat regression module includes both Ordinary Least Squares and Poisson-based regression models, estimated from Maximum Likelihood (MLE) or Markov Chain Monte Carlo (MCMC) algorithms. The current version includes six different models including OLS, Poisson with Linear Dispersion Correction, Poisson-Gamma and a Poisson-Gamma-Conditional Autoregressive (CAR) spatial regression model. The module can handle very large datasets through a Block Sampling approach. There is also a module for applying estimated coefficients to a new dataset to make predictions.

Crime travel demand modeling is a new module in CrimeStat III. It is an application of travel demand modeling, widely used in transportation planning, to crime analysis. The analysis is done by zones. First, crime 'trips' are defined as a link between an offender residence/origin location and a crime location. The number of crimes originating in each zone is counted as is the number of crimes ending in each zone. Second, the model is run sequentially in four separate stages with multiple routine in each stage:

• Trip Generation - Separate models are produced that predict the number of crimes originating in each zone (origins) and the number of crimes ending in each zone (destinations). CrimeStat III uses a multivariate Poisson regression model, with stepwise options, to create the prediction. Trips from outside the study area (external trips) can be added to the origin model to account for travel from outside the region. Once the models are created, a balancing procedure ensures that the number of origins equals the number of destinations.
• Trip Distribution - Using the predicted number of crime trips originating in each zone and the predicted number of trips occurring in each zone, the second stage distributes trips from each zone to every other zone using a gravity model. There are routines for calculating the actual (observed) distribution from individual data, for estimating the prediction coefficients, and for applying the predicted coefficients to the predicted origins and destinations. Another routine allows a comparison of the predicted trip distribution with the observed trip distribution.
• Mode Split - The predicted number of trips for each zone-to-zone pair can be split into likely travel modes using an accessibility function that approximates the utility of one mode relative to the others.
• Network Assignment - Finally, the predicted trips from each zone to every other zone by travel mode are assigned to a likely route based on the shortest path algorithm. The output includes the likely routes taken for each origin-destination zone pair and the total volume of trips on network links. This step requires a travel network, one for each travel mode. There are additional utilities for calculating transit networks from station/stop locations and for testing for one-way streets.

• Parameters can be saved and re-loaded.
• Tab colors can be changed.
• Monte Carlo simulation data can be output.

CrimeStat is accompanied by sample datasets and a manual that gives the background behind the statistics and examples. The manual also discusses applications of CrimeStat developed by other analysts and researchers. The program and sample data sets are in Windows-based zipped files that can be downloaded. The manual is a set of individual chapters in PDF files. They can be viewed online or downloaded. If downloading the PDF chapters separately, they should be saved into the same directory as the CrimeStat program. If the PDF file names are not renamed, they can be accessed directly from the program's help menu.

The CrimeStat Libraries (version 1.0) are component objects that allow for the functions of CrimeStat to be programmed directly into custom software or systems. The CrimeStat Libraries include all of the routines that were developed through version 2.0 of the regular CrimeStat program, including spatial description, hot spot analysis, and kernel density interpolation routines. Additional spatial autocorrelation routines have been included. The libraries can input dbf, shape, and Ascii text files and can output to shape file, MIF/MID files, ASCII text files, and KML files.

The Detection of Crime, Resource Deployment, and Predictors of Success: A Multi-Level Analysis of Closed-Circuit Television (CCTV) in Newark, NJ collection represents the findings of a multi-level analysis of the Newark, New Jersey Police Department's video surveillance system. This collection contains multiple quantitative data files (Datasets 1-14) as well as spatial data files (Dataset 15 and Dataset 16). The overall project was separated into three components:

• Component 1 (Dataset 1, Individual CCTV Detections and Calls-For-Service Data and Dataset 2, Weekly CCTV Detections in Newark Data) evaluates CCTV's ability to increase the "certainty of punishment" in target areas;
• Component 2 (Dataset 3, Overall Crime Incidents Data; Dataset 4, Auto Theft Incidents Data; Dataset 5, Property Crime Incidents Data; Dataset 6, Robbery Incidents Data; Dataset 7, Theft From Auto Incidents Data; Dataset 8, Violent Crime Incidents Data; Dataset 9, Attributes of CCTV Catchment Zones Data; Dataset 10, Attributes of CCTV Camera Viewsheds Data; and Dataset 15, Impact of Micro-Level Features Spatial Data) analyzes the context under which CCTV cameras best deter crime. Micro-level factors were grouped into five categories: environmental features, line-of-sight, camera design and enforcement activity (including both crime and arrests); and
• Component 3 (Dataset 11, Calls-for-service Occurring Within CCTV Scheme Catchment Zones During the Experimental Period Data; Dataset 12, Calls-for-service Occurring Within CCTV Schemes During the Experimental Period Data; Dataset 13, Targeted Surveillances Conducted by the Experimental Operators Data; Dataset 14, Weekly Surveillance Activity Data; and Dataset 16, Randomized Controlled Trial Spatial Data) was a randomized, controlled trial measuring the effects of coupling proactive CCTV monitoring with directed patrol units.

Over 40 separate four-hour tours of duty, an additional camera operator was funded to monitor specific CCTV cameras in Newark. Two patrol units were dedicated solely to the operators and were tasked with exclusively responding to incidents of concern detected on the experimental cameras. Variables included throughout the datasets include police report and incident dates, crime type, disposition code, number of each type of incident that occurred in a viewshed precinct, number of CCTV detections that resulted in any police enforcement, and number of schools, retail stores, bars and public transit within the catchment zone.

This study evaluated CeaseFire, a program of the Chicago Project for Violence Prevention. The evaluation had both outcome and process components.

The outcome evaluation assessed the program's impact on shootings and killings in selected CeaseFire sites. Two types of crime data were compiled by the research team: Time Series Data (Dataset 1) and Shooting Incident Data (Dataset 2). Dataset 1 is comprised of aggregate month/year data on all shooting, gun murder, and persons shot incidents reported to Chicago police for CeaseFire's target beats and matched sets of comparison beats between January 1991 and December 2006, resulting in 1,332 observations. Dataset 2 consists of data on 4,828 shootings that were reported in CeaseFire's targeted police beats and in a matched set of comparison beats for two-year periods before and after the implementation of the program (February 1998 to April 2006).

The process evaluation involved assessing the program's operations and effectiveness. Researchers surveyed three groups of CeaseFire program stakeholders: employees, representatives of collaborating organizations, and clients.

The three sets of employee survey data examine such topics as their level of involvement with clients and CeaseFire activities, their assessments of their clients' problems, and their satisfaction with training and management practices. A total of 154 employees were surveyed: 23 outreach supervisors (Dataset 3), 78 outreach workers (Dataset 4), and 53 violence interrupters (Dataset 5).

The six sets of collaborating organization representatives data examine such topics as their level of familiarity and contact with the CeaseFire program, their opinions of CeaseFire clients, and their assessments of the costs and benefits of being involved with CeaseFire. A total of 230 representatives were surveyed: 20 business representatives (Dataset 6), 45 clergy representatives (Dataset 7), 26 community representatives (Dataset 8), 35 police representatives (Dataset 9), 36 school representatives (Dataset 10), and 68 service organization representatives (Dataset 11).

The Client Survey Data (Dataset 12) examine such topics as clients' involvement in the CeaseFire program, their satisfaction with aspects of life, and their opinions regarding the role of guns in neighborhood life. A total of 297 clients were interviewed.

These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of the files available with this collection and consult the investigator(s) if further information is needed.

This collection was part of a larger two-phase project funded by the National Institute of Justice (NIJ). Phase I focused on the development and estimation of predictive crime models in Shreveport, Louisiana and Chicago, Illinois. Phase II involved the implementation of a prevention model using the predictive model. To evaluate the two predictive policing pilot programs funded by NIJ, RAND evaluated the predictive and preventative models employed by the Shreveport Police Department titled Predictive Intelligence Led Operational Targeting (PILOT). RAND evaluated whether PILOT was associated with a measurable reduction in crime. The data were used to determine whether or not there was a statistically significant reduction in property crime counts in treated districts versus control districts in Shreveport.

The collection includes 1 Excel file (Shreveport_Predictve_Policing_Evaluation_Experiment_Data.xlsx (n=91; 8 variables)) related only to the property crime aspect of the study. Neither data used to perform the outcomes evaluation for the Chicago Police Department experiment nor qualitative data used to help perform the prediction and prevention model evaluations are available.

These data are part of NACJD's Fast Track Release and are distributed as they there received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except of the removal of direct identifiers. Users should refer to the accompany readme file for a brief description of the files available with this collections and consult the investigator(s) if further information is needed.

This study examines municipal crime levels and changes over a nine year time frame, from 2000-2008, in the fifth largest primary Metropolitan Statistical Area (MSA) in the United States, the Philadelphia metropolitan region. Crime levels and crime changes are linked to demographic features of jurisdictions, policing arrangements and coverage levels, and street and public transit network features.

These data are part of NACJD's Fast Track Release and are distributed as they there received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except of the removal of direct identifiers. Users should refer to the accompany readme file for a brief description of the files available with this collections and consult the investigator(s) if further information is needed.

This study was a systematic assessment of the impacts of foreclosures and crime levels on each other, using sophisticated spatial analysis methods, informed by qualitative research on the topic. Using data on foreclosures and crime in District of Columbia and Miami-Dade County, Florida from 2003 to 2011, this study considered the effects of the two phenomena on each other through a dynamic systems approach.

This research sought to examine and evaluate geospatial predictive policing models across the United States. The purpose of this applied research is three-fold: (1) to link theory and appropriate data/measures to the practice of predictive policing; (2) to determine the accuracy of various predictive policing algorithms to include traditional hotspot analyses, regression-based analyses, and data-mining algorithms; and (3) to determine how algorithms perform in a predictive policing process.

Specifically, the research project sought to answer questions such as:

• What are the underlying criminological theories that guide the development of the algorithms and subsequent strategies?
• What data are needed in what capacity and when?
• What types of software and hardware are useful and necessary?
• How does predictive policing "work" in the field? What is the practical utility of it?
• How do we measure the impacts of predictive policing?

The project's primary phases included: (1) employing report card strategies to analyze, review and evaluate available data sources, software and analytic methods; (2) reviewing the literature on predictive tools and predictive strategies; and (3) evaluating how police agencies and researchers tested predictive algorithms and predictive policing processes.

These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of the files available with this collection and consult the investigator(s) if further information is needed.

The study used a place-based method of evaluation and spatial units of analysis to measure the extent to which allocating police patrols to high-risk areas effected the frequency and spatial distribution of new crime events in 5 U.S. cities. High-risk areas were defined using risk terrain modeling methods. Risk terrain modeling, or RTM, is a geospatial method of operationalizing the spatial influence of risk factors to common geographic units.

The collection contains 333 shape files, 8 SPSS files, and 9 Excel files. The shape files include both city level risk factor locations and crime data from police departments. SPSS and Excel files contain output from GIS data used for analysis.

These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of the files available with this collection and consult the investigator(s) if further information is needed.

The study includes data collected with the purpose of determining the geospatial capabilities of the nation's law enforcement agencies (LEAs) with regards to the tools, techniques, and practices used by these agencies.

The collection includes two Excel files. The file "Geospatial Capabilities Survey Data To NACJD V2.xlsx" provides the actual data obtained from the completed surveys (n=311; 314 variables). The other file "Coding Scheme.xlsx" provides a coding scheme to be used with the data.

These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of the files available with this collection and consult the investigator(s) if further information is needed.

This data collection represents an experimental micro-level geospatial crime prevention strategy that attempted to interrupt the near repeat (NR) pattern in residential burglary by creating a NR space-time high risk zone around residential burglaries as they occurred and then using uniformed volunteers to notify residents of their increased risk and provide burglary prevention tips. The research used a randomized controlled trial to test whether high risk zones that received the notification had fewer subsequent burglaries than those that did not. In addition, two surveys were administered to gauge the impact of the program, one of residents of the treatment areas and one of treatment providers.

The collection contains 6 Stata datasets:

1. BCo_FinalData_20180118_Archiving.dta(n = 484, 8 variables)
2. Red_FinalData_20180117_Archiving.dta (n = 268, 8 variables)
3. BCo_FinalDatasetOtherCrime_ForArchiving_v2.dta(n = 484, 8 variables)
4. Redlands_FinalDataSecondary_ForArchiving_v2.dta (n = 266, 8 variables)
5. ResidentSurvey_AllResponses_V1.4_ArchiveCleaned.dta (n = 457, 42 variables)
6. VolunteerSurvey_V1.2_ArchiveCleaned.dta (n = 38, 16 variables)

1. BaltimoreCounty_Bnd.zip
2. BC_NR_HRZs.zip
3. BurglaryAreaMinus800_NoApts.zip
4. Redlands_CityBnd.zip
5. RedlandsNR_HRZs.shp.zip

The purpose of this research was to further understanding of why crime occurs where it does by exploring the spatial etiology of homicides that occurred in Washington, DC, during the 13-year period 1990-2002.

The researchers accessed records from the case management system of the Metropolitan Police, District of Columbia (MPDC) Homicide Division to collect data regarding offenders and victims associated with the homicide cases. Using geographic information systems (GIS) software, the researchers geocoded the addresses of the incident location, the victim's residence, and offender's residence for each homicide case. They then calculated both Euclidean distance and shortest path distance along the streets between each address per case. Upon applying the concept of triad as developed by Block et al. (2004) in order to create a unit of analysis for studying the convergence of victims and offenders in space, the researchers categorized the triads according to the geometry of locations associated with each case. (Dots represented homicides in which the victim and offender both lived in the residence where the homicide occurred; lines represented homicides that occurred in the home of either the victim or the offender; and triangles represented three non-coincident locations: the separate residences of the victim and offender, as well as the location of the homicide incident.) The researchers then classified each triad according to two separate mobility triangle classification schemes: Traditional Mobility, based on shared or disparate social areas, and Distance Mobility, based on relative distance categories between locations. Finally, the researchers classified each triad by the neighborhood associated with the location of the homicide incident, the location of the victim's residence, and the location of the offender's residence.

A total of 3 statistical datasets and 7 geographic information systems (GIS) shapefiles resulted from this study. Note: All datasets exclude open homicide cases. The statistical datasets consist of Offender Characteristics (Dataset 1) with 2,966 cases; Victim Characteristics (Dataset 2) with 2,311 cases; and Triads Data (Dataset 3) with 2,510 cases. The GIS shapefiles have been grouped into a zip file (Dataset 4). Included are point data for homicide locations, offender residences, triads, and victim residences; line data for streets in the District of Columbia, Maryland, and Virginia; and polygon data for neighborhood clusters in the District of Columbia.