PRI staff and faculty associates have substantial experience designing and proposing research designs using a broad range of social science methodology. Proposals developed at PRI have received funding through a variety federal funding agencies, including the National Institutes of Health, the National Science Foundation, and the National Institute of Justice. We also develop proposals for funding through nonprofit foundations and for contracts with local, regional, and state government agencies.

Proposal assistance services at PRI are generally conducted as collaborative projects with SF State faculty members and researchers from other universities, nonprofit institutes, and government agencies. Proposal services are usually offered only for projects in which PRI will materially participate, but unrestricted assistance for SF State faculty is provided subject to resource and staff availability.

back to top

back to top

• Sampling

PRI supports sampling design and sample size calculation, including complex designs: Sampling methods used at PRI include Random Digit Dialing (RDD), listed samples, intercept samples, and hybrid designs. We are currently working to develop expertise in respondent-driven sampling (RDS), an emerging type of chain-referral (snowball) method offering accurate estimates for “hidden populations.”

 

• Computer-Assisted Telephone Interviewing (CATI)

 

   

 

PRI operates its own 25-station telephone survey center for computer-assisted telephone interviewing (CATI) and data entry. The software used (Sawtooth Technologies, WinCATI 4.2) is capable of displaying graphical characters in all languages. The software manages the sample database and guides interviewers and data entry personnel through the questionnaire while automatically compiling a database of responses. The Institute completed a complete upgrade of its telephone facility (computers, workstations, software, and training in 2004, funded by a Major Research Infrastructure grant from the National Science Foundation), and at the same time moved to a larger, modern, and secure office close to the SF State campus.

 

• Internet, Mail, and Mixed-Mode Surveys

In an era of declining response rates and escalating survey costs, it is often useful to consider using multiple modes of survey administration (to increase response rates) or the use of mail and internet methods as a substitute for telephone contact (to reduce cost). PRI regularly administers surveys by mail and internet as the principal contact method as well as in combination with telephone surveying. An advantage of PRI’s WinCATI software is the ability to integrate internet (web-based) data collection with a telephone (CATI) instrument. For example, a telephone survey can be designed so that if a respondent contacted by telephone asks to complete the questionnaire online, the interviewer can enter an electronic mail address and a message with a link to the survey will be issued to the respondent, with the case status recorded to suspend further telephone contact.

 

Some issues to consider in the use of internet, mail, and mixed-mode surveying include:

 

Internet	• Low cost, high convenience, fast implementation. • Response rates will be low without additional contact methods. • Internet-only samples are unlikely to be representative of the general  population, but may be representative of specialized populations.
Mail	• Moderate cost and convenience. • Response rates will be low without additional contact methods. • Time must be allowed for document preparation, mailing, returns, and data entry. • Mail surveys can reach respondents who do not have standard telephone service or internet access. • Large scale mail surveys can be nearly as costly as telephone surveys.
Mixed Mode	• Most convenient for respondents. • Most likely to maximize response rate. • More costly than using a single methodology. • Time must be allowed to prepare multiple versions of instruments, which can be compounded if translations and many revisions are required. • Different survey modes may be subject to different types of response biases.

 

• Personal Interview and Intercept Surveys

  Direct personal interviewing is perhaps the ideal survey method in terms of response rates and data quality, but it can also be costly. PRI conducts a wide variety of projects involving personal recruiting and interviewing, and many of our research staff members have considerable experience. Examples of some recent projects include:

   

  • PRI administered the federal Arrestee Drug Abuse Monitoring program site in the Santa Clara County jail system from 1999 through 2003, completing a rigorous interview and collecting urine specimens from approximately 1,000 respondents per year during quarterly data collection periods.

   

  • In 2002-2003, PRI conducted a major on-board ridership survey on behalf of the Alameda-Contra Costa Transit District, including the development and implementation of a complex sampling plan, obtaining completed questionnaires from more than 15,000 respondents, and preparing comprehensive reports.

   

  • PRI is currently conducting interviews with parents and adolescent children at Kaiser Permanente clinical facilities in four San Francisco Bay Area locations, on behalf of the Prevention Research Center’s “Families Being Together” project designed to prevent alcohol and drug abuse by teens.

   

  For in-person interviewing, PRI uses a Computer-Assisted Personal Interview (CAPI) system for handheld computer technology. We have five AlphaSmart™ Dana devices, which combine the utility of a laptop computer with the convenience of a Personal Digital Assistant (PDA). Combined with the Entryware™ software (Techneos Systems, Inc.) package, we have found this to be a flexible and reliable system for field work.

   

• Multicultural / Multilingual Capabilities

  In a 2003 study by US News and World Report, SF State was ranked among the top ten US universities for campus diversity. PRI has uniquely benefited from this diversity of faculty and students, most recently in our extensive work conducting surveys with the Vietnamese and Latino communities. PRI’s affiliations with a network of faculty, student, research, and community organizations allow us to quickly recruit skilled staff with expertise working in a wide variety of communities. PRI recruits, employs, and provides extensive training to interviewers and supervisors fluent in English, Spanish, Cantonese, Mandarin, Vietnamese, and Tagalog.

   

  For developing and translating culturally sensitive materials, we recommend that sufficient time be planned for steps such as:

   

  • Coordinated teams of translators, with defined roles for primary translation, review, and adjudication with final decision authority

     

  • Independent Back Translation

     

  • Testing and review at each stage by experienced bilingual interviewers

   

  • Use of faculty and external consultants, especially when translating languages not represented among the supervisory staff.

     

  • External expert review for cultural adaptation

     

  • Embedded research to advance methodological understanding

   

  Effective translation procedures must be carefully coordinated with questionnaire design and programming activities, particularly when multiple languages are involved. Because content changes require revisions to translation, CATI programming, and layouts for self-administered materials, it is important to plan for the effective sequencing of these activities.

   

  The development and use of effective procedures for research in multicultural populations is a long term endeavor for PRI and SF State. PRI is committed to the continued identification, development, and advocacy of the best available practices for translation and cultural adaptation.

back to top

back to top

Program evaluation is a systematic way to assess how well a program or intervention is working. Program evaluation is not a methodology in itself, but can entail a number of different research methodologies, both qualitative and quantitative. The reason for conducting a program evaluation is to determine whether a program or intervention is meeting its goals and objectives.

There are two main categories of program evaluation:

• Formative evaluation: Generally conducted prior to the program (needs assessment) or during program implementation (implementation or process evaluation). Needs assessments help to determine whether the program is needed, who needs it, exactly what is needed, and how best to deliver the intervention. Implementation or process evaluations examine the program mid-implementation to determine how well the program is being implemented, what challenges are encountered, what has been learned to date, and where fine-tuning might be useful. The audience for these evaluations may be funders or other stakeholders, but often the primary consumer is program staff looking at how to continue or improve a program.


• Summative evaluation: Often conducted at the end of a program or when a significant phase of the program is completed. Used to determine the eventual outcomes, cost-effectiveness or impact of a program. Includes outcome evaluations, impact evaluations, cost-benefit analysis and cost-effectiveness analysis. Often the audience for this type of evaluation includes funders and other stakeholders as well as program staff.

back to top

PRI frequently uses GIS in the analysis and visualization of research results. When possible, GIS analysis is planned in the research design phase, including the collection of spatial identifiers by which to georeference data. PRI reports often feature thematic maps for visualization of results. PRI employs two staff members skilled in GIS, one of whom is also skilled in cartography. As a unit of SF State, PRI currently has a site license to ArcEditor 9.0 (upgradable to ArcInfo 9.0 if necessary) with upgrade privileges to new releases. PRI uses primarily public domain base maps, such as TIGER files, City of San Francisco geodata, and occasional USGS digital orthophoto quadrangles. PRI has also used client- provided geodata, such as Thomas Brothers’ base maps and SF Department of Public Works proprietary data. Through our connection to SF State’s Institute for Geographic Information Science, PRI has access to additional GIS resources, including a large scale plotter.

back to top

PRI staff members and faculty associates routinely work with a wide variety of analytical tools and techniques,
including:

• SPSS 13.0
• Stata 9.0
• M-Plus 3.01
• Stella 8.1
• Calculation of sampling weights
• Standard errors for complex sample designs
• Qualitative Analysis (Nvivo)
• Power Analysis (NQuery)
• Spatial Analysis (see below)
• Multivariate analysis, including multilevel and structural equation models
• Biostatistical, econometric, and experimental applications
• Secondary analysis of existing data, including US Census and PUMS (Public Use Microdata Sample) data
• Exploratory and graphical techniques