SCPS_OMB_Part_B_-_Supporting_Statement[1]

2. Collection of Information Employing Statistical Methods

Sampling felony case processing in the nation’s 75 most populous counties

The SCPS data collection employs sampling methods. The SCPS 2009 sample was designed and selected by U.S. Census Bureau staff. It is a 2-stage stratified sample, with 40 of the nation’s 75 most populous counties selected at stage one and a systematic sample of state court felony filings within each county selected at stage two. The 75 most populous counties serve as a useful benchmark because they account for about 37% of the U.S. population and, according to the FBI's Uniform Crime Reports, about half of all reported serious violent crimes in the United States. SCPS obtains only one month of felony filing data because of the time and cost constraints inherent in tracking pretrial release and misconduct among felony defendants in state courts.

The first stage of the SCPS sampling design involves selecting 40 of the nation’s 75 most populous counties to participate in the data collection. The 40 counties are selected by stratifying the 75 most populous counties into 4 strata. The strata are based on total population, uniform crime reports (UCR) index crimes, and felony filing numbers. The first stratum includes ten counties all of which are included in the sample with certainty because of their large total populations, heavy felony caseloads, and high index crime numbers. The remaining counties are allocated to the three non-certainty strata based on the variance of felony court filings, index crimes, and total population and are selected at random within stratum. The 40 sampled counties, the total population universe of the nation’s 75 most populous counties, and the sampling weights for the first stage of the SCPS sample are shown below.

SCPS first stage design

SCPS Stratum	Sample of 40 counties	Universe of nation’s 75 most populous counties	First stage sample weights
One	10	10	1.000
Two	7	12	1.714
Three	9	18	2.000
Four	14	35	2.500
Total	40	75	NA

The second-stage sampling (filings) is designed to represent all defendants who had felony cases filed with the court during the month of May 2009. The participating jurisdictions provide data for every felony case filed on selected days during the month of May and the cases are tracked from May 2009 until May 31, 2010. Depending on the first-stage stratum in which it had been placed, each jurisdiction provides filing data for 5, 10, or 20 randomly selected business in May 2009. Data from jurisdictions that are not required to provide a full month of filings are weighted to represent the full month of May. The number of filing days for each of the 4 SCPS Stratum are provided below.

SCPS second-stage design

SCPS Stratum	Number of days of filings provided for the month of May 2009	Second stage sample weights
One	5 days	4.0
Two	10 days	2.0
Three	10 days	2.0
Four	20 days	1.0

Once the data are collected, weights are applied so that the data represent all felony defendants whose cases are filed and processed during the month of May in the nation’s 75 most populous counties in 2009. The un-weighted data represent an estimated 15,000 felony defendants who had a felony charge filed on randomly selected business days in May 2009. When weighted, the data represented an estimated 56,000 defendants charged with a felony offense for the entire of month of May 2009 in the nation’s 75 most populous counties. In addition, Taylor Linearization Methods are used to generate standard error estimates and confidence intervals for the reports that analyze these data.

Data capacity survey

The SCPS sample design described above was constructed in 1988 and has not changed during the various iterations of this project through 2006. The decision to restrict the SCPS sample initially to 40 of the nation’s 75 most populous counties and confine felony filing data to less than a whole month were primarily driven by time and cost restraints and the difficulties inherent in obtaining court case processing, pretrial, and criminal history data. Increases in computing power, combined with growing automation of case management systems and criminal history repositories, has lead BJS to consider redesigning the SCPS sample. Possible revisions could include adding more counties to the 75 most populous county sample, increasing the number of days in May in which felony cases are sampled, and expanding the sample so that estimates could be made about felony case processing outside the nation’s 75 most populous counties.¹

BJS conducted a number of activities in 2008 to explore various redesign options for the SCPS court processing data collection. In partnership with the Pretrial Justice Institute and the Urban Institute, we explored a number of different sample designs using historical SCPS data collected from counties representative of the nation’s 75 most populous and other public data sets that may help identify bias in the SCPS sampling design. With this limited information, we determined that the current SCPS model was achieving its stated goals as well as possible within its current resource parameters. We found no biases associated with seasonality, the manner in which the counties and filing days were drawn, or the amount of time the cases were followed. However, the initial SCPS redesign activities were only able estimate the universe of felony cases filed based on data collected from past SCPS data collections. We had no basis for understanding level of felony case filings in state courts from a more nationally representative sample.

In order to determine the feasibility of enhancing the 75 most populous county sample, to improve the coverage, stratification and selection process of counties and records, including increasing the number of sampling days, and expanding the sample outside the nation’s 75 most populous counties, SCPS will survey the automation levels of court case processing systems in two groups of courts. The first group will involve obtaining automation information for courts in the entire universe of the nation’s 75 most populous counties. The second group will entail gaining automation information for courts located outside the 75 most populous counties but within the nation’s 900 most populous counties. A total of 125 courts will be surveyed in second group. The 125 courts will be drawn by stratifying the 825 counties (900 most populous counties subtract 75 most populous counties) by census region, levels of urbanization, population size, and UCR index crimes.

In conclusion, a total of 200 courts will be surveyed to assess how the different information components in SCPS are managed (whether the different information items are managed by one source or by separate offices), their overall levels of data automation and the level of automation for each of the types of information desired, data storage capabilities, and capacity to provide electronic data for the SCPS data collection. The information gleaned from this survey will be used to evaluate the costs, time, and overall feasibility associated with expanding the SCPS sample for future data collections.²

2. Procedures for Collecting Information

Before individual felony case data are collected, sample lists will be drawn from each of the selected 40 counties. The sample lists will consist of all felony cases filed in the selected counties during certain randomly selected days in May 2009.

Once the sample lists are generated, data on felony cases filed in the 40 selected counties will be collected from a variety of sources. These will include court administrative files, pretrial case management systems, jail records, prosecutor offices, state criminal history record files, and other statistical agencies. Each participating court in the SCPS project will be contacted (see attached contact letter) for the purpose of locating and hiring a contractor to either manually complete each data collection form or to write software capable of generating electronic downloads of SCPS data. Contractors will also be hired to complete the court automation survey forms. All contractors will be compensated according to the number of hours worked completing the data collection forms and court surveys.

All data collection contractors will undergo extensive training and will be managed by staff from the data collection agent. When collecting data that entails the coding of administrative court records, prosecutor files, jail information systems, or criminal history rap sheets, it is not uncommon for respondents to code responses that might contain logical mistakes or errors. In order to prevent these errors, the data collection agent reviews all data collection forms submitted and it’s not unusual for the agent to make several, on the order of 3 – 4 contacts, with the respondent to resolve these errors. Once the respondent has become familiar with the data collection procedures, fewer contacts are needed.

Typically, the advance letter is sent to the chief administrative official of the local court or pretrial diversion program. Sometimes these chief administrative officials have authority to decide on county participation; in other instances, initial requests are sent to the chief trial court judge who then determines whether the county will participate. Once the chief justice or administrative court/pretrial official offers approval, we are given contact information for the individual collector or systems administrator who provides direct assistance to this project. Respondents were identified from a list of participants who assisted in the 2006 SCPS data collection.

3. Methods to Maximize Response Rates

In order to maximize the response rate and minimize non – response bias, every attempt will be made to collect complete and accurate information on all 15,000 felony defendants. The data collection agent will, if necessary, travel to sites that are unwilling to participate in the SCPS project. For the 2006 SCPS project, information was collected on every felony case processed in the nation’s 75 most populous counties. Moreover, the amount of missing and incomplete information was negligible and did not hinder this project. We anticipate similar response patterns for the 2009 SCPS project.

4. Testing of Procedures

The SCPS data collection forms were pretested prior to data collection. Several respondents completed the data forms and provided feedback in terms of clarity and accuracy. In addition, the SCPS redesign project involved consultations on the overall substance of information collected and the format in which questions are asked on the data collection form. Appropriate revisions and modifications were made to these data forms based on this feedback. Some of the revisions included obtaining more detail information on the relationship between commercial bail bondsmen and pretrial release agencies and delving into the utilization of immigration holds as a means of keeping felony defendants detained during the pretrial process.

5. Contacts for Statistical Aspects and Data Collection

The prosecution and adjudications staff, at the Bureau of Justice Statistics, take responsibility for the overall design and management of the SCPS data collection, including the development of the questionnaires and the analysis and publication of the data. BJS contacts include

Duren Banks, Chief

Prosecution and Adjudications Statistics Unit

Bureau of Justice Statistics

810 7^th Street, NW

Washington, D.C. 20531

(202) 307 – 0765

Thomas H. Cohen, Statistician

Prosecution and Adjudications Statistics Unit

Bureau of Justice Statistics

810 7^th Street, NW

Washington, D.C. 20531

(202) 514 – 8344

Tracey Kyckelhahn, Statistician

Prosecution and Adjudications Statistics Unit

Bureau of Justice Statistics

810 7^th Street, NW

Washington, D.C. 20531

(202) 353 – 7381

b. Persons consulted on statistical methodology:

Carma Hogue

Chief, Statistical Support and Consulting Staff

Economic Statistical Methods

and Programming Division

U.S. Census Bureau

Washington, DC 20233

(301) 763 – 4882

Robert Santos

Senior Institute Methodologist

The Urban Institute

2100 M Street, NW

Washington, DC 20037

(202) 261 – 5904

c. Persons consulted on data collection and analysis:

Gene A. Lauver

Supervisor of Grants Administration

Regional Justice Information Service (REJIS)

4255 West Pine Blvd
St. Louis, MO 63108
(314) 535 – 1950

1 Possible designs include samples that could provide estimates of felony case processing in the nation’s 100, 150, 450, and 900 most populous counties. These estimates are being considered because they met certain population benchmarks. Forty – two percent of the U.S. population resides in the nation’100 most populous counties, while 51% and 73% of the nation’s population resides in the top 150 and 450 counties, respectively. About 85% of the U.S. population resides in the nation’s 900 most populous counties.

2 The 2009 SCPS sample will remain unchanged from SCPS prior projects. Information gained from the court survey will be used to inform sampling stratagems for future SCPS data collections.

5