- The CATI system eliminates the need for paper forms and reduces the responsibility of the interviewer to monitor the order in which questions are asked. This allows for easy administration of survey instruments with conditional branching, as well as items which require a random-order administration sequence.
- The system eliminates the chances of missed questions and will not accept any value outside the range of acceptable responses. These features substantially reduce the possibility for interviewer and coder errors.
- The system has an automated sample management program which eliminates the need for supervisory staff to maintain the status of each case in the sample.
- The system provides on-screen display of instructions for interviewers as required, and on-screen "fill" display options for items with variable values and ranges.
- The system allows for the entry of open-ended responses, verbatim text and notes.
- The system has the ability to interrupt an interview in mid-course and resume the interview at a later time.
- The system has the ability to move backward and forward to change prior and subsequent entries.
- Responses to questions are keypunched into an electronic database as interviews are conducted, eliminating the need for subsequent data entry.
- The data are available for analysis immediately upon completion of each project.
Sampling Procedures
The Lab uses the Genesys Sampling System to generate random telephone number samples. The system uses a list of all possible telephone numbers in the US to randomly generate telephone numbers.
With the Genesys system, we can:
- target populations by selected demographic characteristics
- analyze the neighborhood characteristics of participants
- target geographic areas using census tracts
Internet Surveys
The PPRL can use the Internet to collect survey information, evaluate training programs, and monitor client satisfaction. The PPRL has experience in designing and conducting web-based surveys for a variety of clients, populations, and situations. We can work with you in developing sampling strategies, designing instruments, posting web surveys and forms, and quickly analyzing the survey results. Web-based surveys and forms can also be used in "mixed mode" situations where telephone, fax, or mail contact options are also used.
Mail Surveys
The PPRL offers complete mail survey services. From the questionnaire design to the final report, the PPRL can handle all your survey needs. Superb respondent tracking tools allow us to keep track of the status of each respondent case throughout all points in the research project. This allows us to give clients up-to-date project status reports as frequently as they need. We can conduct all phases of the mail survey process, from questionnaire design, and formatting, to mail out and log in, data entry, data delivery, and report writing.
In-Person Surveys
For selected projects, PPRL may be able to offer in-person interviewing. Projects for which this option would make sense include intercept interviews with clients at their point of service and small neighborhood surveys.
Mixed-Mode Surveys
Often the best method is a combination of techniques. The PPRL has extensive experience in mixed-mode surveys that combine multiple methods (e.g. mail, email and telephone).
Focus Groups
PPRL has expertise in conducting focus groups, both as a primary data collection mechanism and as a step in the questionnaire development process.
Scanning System
PPRL's Scanning System (data capture) helps in scan, recognize, verify and export data and images quickly, accurately and cost-effectively. Our system is capable of processing thousands of surveys every day.
Our software is capable of reading hand printed (pen or pencil, any color) alpha and numeric data, such as a name, address and phone number, as well as common check boxes. Data is automatically updated into a database or spreadsheet as required.
Using scanning technology means we reduce manual data entry considerably, thereby reducing staff costs and increasing the speed of results. Data can be saved as any of the Microsoft Office suite, such as Excel or Access, csv files, or many complex data formats.
Consulting
The PPRL provides consulting on all aspects of Survey Research projects to the LSU community and others. Please contact us about consulting for:
- Questionnaire Design
- Research Design
- Sampling
- Secondary Data Analysis
- Statistical Analysis
- Data Entry and Database Management
Quality Control
As a research institution the LSU Public Policy Research Lab strives to provide the highest quality work. To do this several procedures have been put in place to review the quality of the data gathered.
Of course, any research involving people carries the possibility of unforeseen problems. However, the following are designed to maximize the chances of a smooth project:
CATI
- Computer Assisted Telephone Interviewing (CATI) combines interviews with data entry. By eliminating the data entry step, a potential source of error is eliminated. In addition, the CATI software can be programmed as to only allow a particular range of answers, reducing keystroke error.
Pretesting
- All surveys are pretested to remove wording errors and programming problems. During the pretest, the validity of the questions is verified. If respondents do not understand the questions, alterations can be made for the full survey. While all surveys are pretested, particularly large or complex surveys may be tested for an extended period of time.
Supervisor Observation
- Every thirty minutes a supervisor listens to a random surveyor conduct a survey using the Supervisor Observer feature. This allows the supervisor to listen to a phone conversation without being heard. The supervisor records the time of survey as well as comments that are shared with the surveyor as well as the lab manager.
Midpoint Data Check
- For a typical survey, the frequencies to all responses are checked approximately half way through its completion. For larger or more complex surveys, this can happen more frequently. This process gives the principal investigator a window into the data collection. Adjustments in the sampling procedure or survey instrument can then be made as appropriate.
Survey Verification
- Small sub samples of surveys are recalled to verify the data collection process. The supervisors will check to determine if the survey was given to the correct person and if the surveyor was professional and courteous.
|
