Making Survey Research Smoother: A Guide for Principal Investigators
The SRC is focused on high quality research. This page describes the functions of the SRC and practices researchers can follow to help their project go more smoothly.
Project Elements
The SRC handles both small and large projects. SRC services include:
- study design
- sampling
- project management
- survey data collection
- data delivery
- data analysis
Setting up a Project
New Project Budgeting & Design, contact:
- Kurt Johnson - Director kdj11@psu.edu
- Patricia Wamboldt - Assistant Director pmw101@psu.edu
Existing project extension and budgeting, contact:
- Stacey Davis - Budget Coordinator sld34@psu.edu
Initiating a Project
Since the implementation of a research project involves numerous SRC staff who are engaged in multiple projects at any one time, the following must occur prior to project implementation:
-
Up to 4 weeks prior to beginning to set up your project, the SRC must be
given the IRB approval number (for projects involving human subjects).
In some cases the SRC can begin work without this approval, but these exceptions
must be explicitly arranged through the Office of Research Protections.
- Up to 4 weeks prior to beginning to set up your project, the SRC must be given the contact person in your college or organization who will certify the funds to match the approved budget are available for disbursement.
It is important to keep the SRC apprised of the progress towards receiving a grant if this budget projection is used as part of a grant proposal. This is especially true for projects that will require significant SRC resources. For example, telephone studies may require advance notice of 6 months or longer depending upon the number of CATI stations committed to other studies. (Also, it is helpful for us to know if a project that we budgeted is not funded.)
The SRC will provide a detailed budget estimate based on the information provided.
Please notify the SRC about any project changes that could affect the budget (i.e. project grant not fully funded). Any modifications of the study design, instrument, mode, etc. will require a re-negotiation of project timelines and proposed budget.
If you are interested in using the SRC to provide these services, please respond to the SRC's budget estimate via email accepting the budgeted amount and providing the following information:
- Main Contact:
- PI:
- Person in your College to Verify Funds:
- College:
- Department:
- Title of Project:
- Grant Project Period:
- What date should the SRC start your project:
- What date should the SRC complete your project:
- Sponsor:
- SRC's Budget Amount for your project: $
- Person to Invoice:
- Your Project's Budget number:
- Your Project's Fund number:
- Are there human subjects? Yes or No
- If yes, we need IRB number:
- If this is a design phase project and no IRB number has been provided, please give brief explanation to be included on request for a fund number:
Beginning Work on the Project
When you notify the SRC of acceptance of the budget and provide the information above, please notify your college contact that the SRC will be contacting them for approval. College research coordinators are:
- LA-Trish Alexander (Alexander@psu.edu)
- HHD-Boe Itinger (jri1@psu.edu)
- ED-Vicki Spadaccio (vls4@psu.edu)
- AG-Judy McCormick (judymc@psu.edu)
It takes at least a week to set up your account.
Required Documents and Timeline
Once the account is set up, the SRC can begin working on your project. At that time you will meet with an SRC Project manager to initiate the SRC's work on your project (at that time provide IRB approval and all necessary documents such as survey instrument, letters, etc.).
The following documents are usually required for the SRC to facilitate your project:
- Study Description/ Proposal
- Grant Proposal (if applicable)
- IRB application, approval and updates
- Survey Questions
- Agreed upon budget for the SRC's work
The actual timeline for completing this project depends on its complexity, the quality of the sample provided, and any alterations made after we begin working on your project which delay project preparations and implementation.
Checklists and Tips
Initial Project Action:
Depending on the project, some of the questions and suggestions may not apply.
Communicate to the SRC your project's details:
- What is the purpose of the study?
- What is the project timeline?
- What mode(s) of data collection will the study require?
- What type of sample will the study require (what is the target population)?
- How many cases do you hope to acquire for the analysis?
- How will the sample be obtained?
- Will the study be piloted? In what way(s)?
- Will the study use participant incentives?
Communicate to the SRC your project's funding status:
- Has the study gained departmental/college financial approval? If not, when?
- How will the project be funded?
Communicate to the SRC your project's Institutional Review Board status:
- Have you checked on IRB requirements? http://www.research.psu.edu/orp/
- Does the study require assistance from the SRC to gain project approval?
- Does the SRC have a copy of the project's approval letters and/ or IRB number? (Make sure the SRC is made aware of all modifications.)
Tips for working with an SRC Project Manager (PM):
Timeline/ Deadline Considerations:
- Negotiate a realistic timeline with the Project Manager (PM) for specific task completions. Substantial changes could extend deadlines.
Survey Questionnaire Design Considerations:
- Ask the PM about special IRB considerations.
- If needed, the SRC website has tutorials for setting up web and paper surveys that may save the project time and cost. (www.ssri.psu.edu/survey)
- Provide the most complete version of the survey to the PM. He or she can help you edit the survey instrument. Only when it is communicated to the PM that the survey is in its final version will it be ready for programming. Any changes after the final version will effect the project timeline!
- If changes have to be made to the survey, present changes (track them) as clear as possible so PM knows what is being changed. If there is more than one investigator all of the PIs should agree on survey changes before giving these changes to the PM. Any changes to the survey instrument must be sent to the IRB.
Mail Survey Considerations:
- Ask the PM about special IRB considerations.
- Consider time and cost of using open-ended questions as the responses will have to be typed in.
- Make sure all addresses are formatted the same (in Excel format). Please provide the whole address.
- Allow ample time for mail collection since odds and ends cannot be scanned once data is cleaned (late surveys will be returned to PI).
- Allow ample time for scanning: the length of the survey determines how long the scanning process will take.
Telephone Survey Considerations:
- Ask the PM about special IRB considerations.
- Be sure the survey makes sense when read aloud (survey format different).
- Decide if you would like the SRC to screen the sample for a specific target population (could affect timeline).
- Consider that the length of survey could vary by the individuals surveyed (affects cost). Test the length of the survey by reading the interview out loud and allowing time for response.
Field Study Considerations:
- Ask the PM about special IRB considerations.
- Check with the PM to find out what type of equipment is needed.
- Clarify who is recruiting the participants
- For school studies, obtain class lists, labels, parent names, phone numbers, addresses (electronic list in excel is very helpful).
- Determine a data coding system for each wave or form; provide coding system if possible.
Web Survey Considerations:
- Ask the PM about special IRB considerations.
- If providing the sample, the SRC must receive valid email addresses.
- Test the survey several times and communicate any programming errors to the PM.
Multi-Modal Project Considerations:
- Ask the PM about special IRB considerations.
- As much as possible, keep the same survey contents and formats for multi-modes.
- Consider how the data will be merged.
Data Considerations:
- Ask the PM about special IRB considerations.
- The data cannot be provided to you until it has been cleaned (for example, verifying information and correcting any errors or missing data issues). Please allow adequate time for cleaning (some modes take more time than others).
- If the data will be analyzed by the SRC, please let the PM know the hypotheses you will be testing, and decide what sort of analysis your project requires, such as: descriptives, cross-tabulations, regression, etc.
- Let the PM know what format you would like the data.
Quality Control Procedures
The following information includes brief summaries of SRC quality control procedures for each mode of survey research. We welcome any input from Principal Investigators. Please let us know your requirements as we are always willing to accommodate specific requests.
Self-administered Surveys
Completed, returned surveys are scanned using TeleForm software. Data is exported into an SPSS file and examined for problems using descriptive and frequency runs. Problems are identified and corrected. A sample of surveys is checked manually to ensure the scanning process was accurate.
A "verification" log is maintained during the scanning process noting any illegible, multiple response issues, or other specific problems. This log is given to the PI as part of the documentation provided for the study. Images of each survey are created during scanning, and also provided to the PI. A second SRC staff member reviews the cleaned data set and recodes any labeling. The PI is also supplied with a codebook, a short methodology write-up, as well as copies of the survey and all letters or other correspondence that was included in mailings.
Telephone Surveys
Interviewers use a Computer Assisted Telephone Interview software called WinCATI to complete surveys over the telephone. The CATI software automatically takes the interview through any skip or branching patterns within the survey, eliminating the need for separate data entry and allows for frequent tabulation of the data as the survey proceeds. The CATI software also gives the interviewers the flexibility to go back and correct answers or add respondent comments to any question. The verbatim responses and the respondent's comments are stored in a text file that can easily be sorted by question in order to develop coding categories.
Phone interviewers attended training sessions and practiced interview techniques with a supervisor. Quality control is managed through monitoring the interview process using visual and audio monitoring. At least 10% of completed interviews are verified by calling the respondent back to assure the interview took place.
Web Surveys
Electronic surveys can be either emailed to respondents or set up on a website that the respondent signs into using a password. Web surveys are run on a firewalled web server with 128-bit SSL security. Surveys are stored in a SQL database which can only be accessed locally. ID numbers are generated for each survey, and surveys can either be locked down so that respondents may only take a survey one at a time or can be allowed to go back and review their answers. All web data is backed up nightly. All answers are kept confidential and only aggregate results are reported.
Field Studies (In-school or personal interviews)
Using scannable or computer software for field data collection increases data quality. Self-administered surveys use TeleForm, a survey design and scanning software. The SRC can produce pencil and paper surveys that are automatically scanned into data files. Face-to-face interviews are conducted by an interviewer who is physically sitting with the respondent. Data are collected using scannable surveys, a computer program similar to WinCATI, video cameras, and saliva collection.
Standard quality control procedures include:
- After training, in-home interviewers must submit a practice tape for review or conduct an in-home interview with the project manager present before they are allowed to do an interview on their own.
- New interviewers have all tapes quality-controlled until they reach criterion levels on 3 tapes, then tapes are chosen periodically to check quality. Every tape is briefly viewed for light and sound quality as it is copied for storage.
- Paper surveys are check by each interviewer for completeness and quality-control periodically by central staff.
- In school surveys are checked extensively to make sure none are left undone or left behind.
- Follow-up calls are periodically conducted to ask respondents (in home or in school) about their experience.
Last modified: 10/24/07 | Contact Webmaster
