B. Collection of Information Employing Statistical Methods
1. Universe and Respondents Selection
1.1. Parole Districts and Field Offices
The frame for sample selection will be provided by the 2006 Census of State Parole Offices. The Census will include aggregate-level data on the number and characteristics of those under post-release supervision, by state, regional, administrative, and field offices. The universe for the FPS will include all parole offices in the Census; preliminary estimates include 1,804 individual field offices and 562,611 parolees.
A simulated sampling strategy was developed for the national survey plan. This strategy may be slightly altered pending actual Census results (e.g., exact number of certainty districts may be altered). However, the intended strategy involves a multi-stage selection. At the first stage a nationally representative sample of 100 parole districts (administrative level above the field office) will be selected with probability proportional-to-size (PPS), with each of the district total active client estimates used as the measure-of-size (MOS).
Based on NORC’s sampling simulations, it is estimated that approximately 300 districts will be included in the final district level frame. This frame will be partitioned into two parts – certainty districts and non-certainty districts. Based on the simulation, it is estimated that approximately 41 districts (with the highest parolee counts) will be selected into the sample with certainty. Approximately 59 additional districts will then be selected PPS from the remaining districts, for a total of 100 districts selected.
Prior to selection, small districts will be collapsed with adjacent districts to allow for a sufficient sample size to complete 65 by sampling, at most, half of the initially estimated eligible parolees. Also, non-certainty districts will be ordered using a serpentine sorting technique which will maximize the stratification effects and ensure a) a geographic spread of districts among census divisions and states within census divisions, as well as, b) a spread of districts within states based on district size.
NORC will contact each state to obtain permission to survey within the state and to obtain a listing of all offices within the chosen districts. The number of clients under active supervision at each office on a selected date will be obtained. Cooperation from districts is expected to be very high. However, if a district denies cooperation, it will be replaced with a random selection from a similar geographic area.
Field offices will be collapsed into geographic groupings using the same specifications used to collapse districts. From one to seven field office groups will be selected from each district, again using PPS sampling. The final sample, based on actual numbers of districts, offices, and active clients, is likely to vary. However based on earlier simulations, it is estimated that 200 field office grouping will be included, with 60 involving the largest offices and 140 field office groups with smaller sizes.
Parolee Selection
Each office included in the sample will be asked to provide a gender-identified roster of former prisoners under active parole supervision. All eligible parolees must be 18 years of age or older. Individuals who are currently incarcerated, have absconded, have a warrant out for their arrest, or have completed parole, or have been transferred to another office are ineligible.
Prior to parolee selection, the roster of parolees in each field office of the selected grouping will be checked for completeness and eligibility. The site set-up staff will select approximately 130 or 260 parolees depending on whether 65 or 130 completed interviews are expected. Having identified the gender of each rostered parolee, female parolees will be oversampled, to the extent possible, to provide for similar precision in results. In addition, a 20 percent reserve sample will be held to allow for differences between estimated and actual numbers of completes.
Final former prisoner completed interviews are anticipated at approximately 16,500. Based on the census received, it is expected that a sample of 33,000 parolees will be needed. Sample sizes by office will be adjusted as needed to obtain the projected number of completes. An additional sample of 6,000 rostered parolees will be held in reserve in the event of an imminent short fall. The NORC statistician will monitor sampling assumptions throughout the data collection period, including eligibility and completion rates, and release of reserve sample.
2. Procedures for Information Collection
Data collection procedures include computerized interviewer-administered interviews and Audio Computer-Assisted Self-Interviews.
The methods proposed for use in data collection are as follows:
State/District Recruitment
Once sampling has been completed, state and district level administrators will be approached to obtain approval. Each sampled parole office will be contacted to solicit participation. A contact person will be designated at each facility.
b. Sampling of Parolees
Approximately 4-5 weeks prior to data collection, the parole office will provide a roster of parolees under active supervision who are 18 and older , and who have not absconded, been reincarcerated, been transferred, or had a warrant issued. A random sample of parolees will be drawn from the roster.
c. Contact and Background Information Data Collection
A site set-up team will supply the parole office with the list of sampled cases. They will be asked to provide contact and background information on the sample. The site set up staff will work with the parole office (and state agencies) to determine the most efficient manner for collection of this information. If automated data is available, it may be provided in that manner. If not, a records form (FPS-1, see Attachment D) will be provided to be completed in paper-and-pencil format, or through a web entry screen. If the parole office agrees to provide access to the data, field staff from the contractor could also enter the data electronically.
Contact information will be used to set appointments for the interviews with the sampled parolees.
Administrative records data will allow researchers to compare demographic characteristics of responding parolees with those who did not participate, in order to determine if there is any bias introduced from non-respondents.
d. Appointment Setting
Approximately two weeks prior to the interview period, a letter inviting participation will be mailed to each sampled parolee. In addition, staff from the contractor will begin to call all sampled parolees to set appointments. After the interview period begins, interviewing staff will continue to call sampled parolees to set appointments or to reschedule missed appointments.
e. Interviewing at Parole Offices
Interviews will take place at each parole office for an approximate two week period. Two to four interviewers will be available, depending on office size, to complete the interviewing process. Interviewers will follow the protocol determined with each office as to parole office entry. In most sites during the pretest, parolees checked in at a front desk and interviewers were buzzed to meet them in the waiting area.
Interviews will take place in a private interviewing area. The interviewer will read a consent form to the former prisoner and solicit his/her participation. If the respondent consents, the interviewer will begin administering a brief set of demographic questions that includes age, race, ethnicity, length of incarceration, and reason for incarceration. After completing the demographic section, the respondent will receive a brief tutorial on answering questions on the touch screen computer, and will begin the more sensitive sections of the interview in complete privacy. In order to allow parolees with reading difficulties to participate, the respondent will wear a set of headphones and hear the questions being read as they appear on the screen. The former prisoner will enter his response by touching a button on the screen – no computer expertise is required. Neither the interviewer nor any parole office staff will be aware of the parolees’ responses. At the end of the questionnaire, the respondent will turn the computer back to the interview, receive the incentive, and be escorted back to the waiting area. The interviewer will then finish the process by answering a set of debriefing questions about the interview.
3. Methods to Maximize Response
Every effort is being made to make the survey materials clear and simple to use. The confidential nature of the data collected is clearly and repeatedly explained in the consent process. The FPS questionnaire has been designed to maximize respondent comprehension and participation and minimize burden. Some examples include an easy to use touch-screen interface with the questions simultaneously delivered via headphones. A Spanish version of the questionnaire will be available for non-English, Spanish speaking respondents. Field staff from the contractor will be available to answer any questions that respondents may have, including bilingual staff who can answer questions in Spanish. A flyer with a listing of local counseling services will be provided to all respondents, with telephone numbers listed for those interested in obtaining counseling services or assistance following the survey.
The greatest difficulty in obtaining response for the parolee population is locating the respondent to invite participation. During the pretest, parolees were often found to have unstable residences and/or non-working phones. The letter mailed to each respondent, providing a toll-free number increased response in this population. Often the mailing address used by the parole office was a parent or family member who could see that the respondent received it. This letter both increased the number of call-ins for appointments and facilitated response when calls were made to them, as they were familiar with the project. The $50 incentive was also instrumental in increasing response among this population.
During the pretest, some respondents were unable to get to the parole office during the scheduled interviewing period due to work obligations. Staff will work with each office to determine if some late afternoon and evening hours can be available for interviewing, with reimbursement for parole office staff who will keep the office open after hours.
4. Test of Procedures or Methods
The interview and data collection procedures were tested in a pilot study conducted Oct. – Dec. 2006, results of which have been cited throughout the supporting statement.
5. Consultation Information
The Corrections Statistics Unit at BJS takes responsibility for the overall design and management of the activities described in this submission, including sampling procedures, development of the questionnaires, and the analysis of the data. BJS contacts include:
Christopher Mumola
Corrections Statistics Unit
Bureau of Justice Statistics
810 Seventh St., N.W.
Washington, DC 20531
(202) 353-2132
Allen J. Beck, Ph.D.
Principal Deputy Director
Bureau of Justice Statistics
810 Seventh St., N.W.
Washington, DC 20531
(202) 616-3277
The Project Director is:
Candace M. Johnson
Senior Research Scientist
NORC
1350 Connecticut Ave., NE, Suite 500
Washington, DC 20036
(202) 223-7936
| File Type | application/msword |
| Author | Scarbora |
| Last Modified By | Scarbora |
| File Modified | 2007-08-24 |
| File Created | 2007-08-24 |