The Journal of Extension - www.joe.org

December 2017 // Volume 55 // Number 6 // Tools of the Trade // 6TOT3

Using Survey IDs to Enhance Survey Research and Administration

Abstract
Survey IDs are short strings of unique characters assigned to each recipient in a sample population. Extension research can benefit from the improved organization of survey implementation and data collection, better researcher-respondent communication, and reduced survey material costs supported through the use of survey IDs. This article outlines how survey IDs can provide a more efficient approach to survey administration and data management and includes suggestions about overcoming some limitations of survey IDs. Best practices for creating and using survey IDs when organizing and administering survey research also are suggested.


Catrin M. Edgeley
Ph.D. Candidate
Department of Natural Resources and Society
College of Natural Resources
University of Idaho
Moscow, Idaho
cedgeley@uidaho.edu

Introduction

Survey IDs (also known as unique identifiers, participant IDs, or geocodes) are short strings of characters assigned to each questionnaire recipient in a survey sample population. Extension professionals can use survey IDs when conducting research to improve communication, reduce costs, and increase organization, thereby streamlining the survey administration process for both themselves and respondents. Using IDs in survey research and administration is common, but there is a shortage of information about effective approaches to doing so. Appropriate use of survey IDs, benefits and challenges to their implementation, and recommendations regarding their application are provided herein and are informed by the experience of administering a mixed-mode survey of approximately 2,700 households in Valley County, Idaho.

When Are Survey IDs Appropriate?

Survey IDs are often beneficial when one or more of the following circumstances exist:

  • The survey administration approach involves efforts to contact each potential respondent multiple times.
  • The researcher has a limited budget for an administrative approach that involves mixed-mode implementation or multiple iterations (e.g., reminder letters mailed strategically after a questionnaire has been provided).
  • Survey participants are selected using probabilistic sampling.
  • Survey data will be integrated with existing data (e.g., spatial information).

Benefits and Challenges of Using Survey IDs

Benefits

Benefits of using survey IDs relate to communication, cost reduction, and data management.

  • Improved communication. Survey IDs can help alleviate challenges regarding researcher-respondent communication. Creating a database of survey participants' information paired with their respective survey IDs is useful for efficiently responding to participants' inquiries, such as requests for replacement questionnaires, or confirming questionnaire receipt.
  • Reduced costs. Recording the survey ID associated with each completed questionnaire as it is received can reduce printing and mailing costs for administration approaches that involve providing a sequence of materials to each recipient. Identifying survey recipients who have yet to respond and sending reminders to only those people saves time and money and ensures that those who have already returned completed questionnaires will not receive follow-up materials that are no longer relevant to them. Tracking survey response on the basis of IDs also helps researchers investigate nonresponse bias.
  • Minimized response rate errors. By assigning each survey recipient a unique survey ID, researchers can track accidental or repeated data submission attempts in both online and paper surveys, ensuring that each respondent is counted only once. Survey IDs also provide a further validity check on responses by ensuring that only data from those in the intended sample are accounted for in the final data set. Requiring survey ID entry at the beginning of online surveys can minimize survey response from unintended participants, particularly when publicly available web links are used (Baatard, 2012).
  • Streamlined data integration. Assigning survey IDs can help researchers link completed survey responses with other existing data (e.g., demographic information or spatial attributes associated with a recipient's address) or forthcoming data. Using consistent survey IDs is particularly beneficial for organizing and merging data collected via longitudinal approaches that survey the same respondents multiple times, such as the Delphi method (Gamon, 1991).
  • Improved sample data management. Using survey IDs to differentiate respondents (instead of using personal information such as phone numbers or addresses) can prevent confusion if respondent contact information is incorrect or changes during survey administration (Dillman, Smyth, & Christian, 2014).

Challenges

Challenges of using survey IDs center on the need for transparency, attention to the potential for human error, and awareness of associated costs.

  • Distrust. The presence of survey IDs can create distrust among some survey recipients, predominantly due to concerns about the anonymity of survey responses. Providing a transparent explanation in supporting survey materials (e.g., a cover letter) about the purpose of survey IDs can reassure respondents and help researchers gain their trust (Dillman et al., 2014). It is useful to briefly outline steps taken to ensure respondent confidentiality (e.g., "Survey IDs will be destroyed after data collection, and your personal information will not be associated with your responses in any way") (Cui, 2003).
  • Human error in online surveys. Requiring respondents to enter a predetermined unique survey ID at the beginning of an online survey may increase the likelihood of mistakes or typos, potentially leading to responses being incorrectly linked with existing data, associated with an incorrect ID, or incorrectly organized (Baatard, 2012). Searching completed data sets for duplicated or incomplete survey IDs can help identify these errors.
  • Potential costs associated with personalization of survey materials. Printing unique IDs on paper survey materials can increase administration costs. Printing stickers and affixing them to survey materials instead is one possible alternative for reducing expenses. Costs also can be reduced by including survey IDs only when necessary—generic materials such as presurvey introductory letters may not require this information.

Creating and Using Survey IDs

What Should Survey IDs Look Like?

Survey IDs are typically presented as an easily identifiable string of characters, comprised of numbers or a combination of numbers and letters. Short numbers can be effective for smaller survey populations. Using a letter at the beginning of a survey ID (e.g., X02657) or hyphenation (e.g., 124-387) allows for quick identification. Potentially ambiguous characters (e.g., the number 0 versus the letter o) should be avoided (Couper, Traugott, & Lamias, 2001).

How Should Survey IDs Be Assigned?

Researchers should adopt a logical, systematic approach to assigning survey IDs. This approach could be based on alphabetical order, distribution methods, or predetermined groups (e.g., survey respondents in one county might have IDs that begin with X, and another county may be assigned Y). It is useful to create a master file in Excel populated with sample data and then assign each potential survey recipient an ID in a separate column. Databases that record assigned survey IDs also can be created using GIS programs if IDs are allocated on the basis of spatial data.

Where Should the ID Be Placed on Survey Materials?

Placing survey IDs in plain view is essential to maintaining transparency. Common ID placements include the front or back cover of a paper survey, the top of a corresponding letter or postcard, or the beginning of an email. Some participants may attempt to remove survey IDs on paper surveys over concerns about the anonymity of their responses. Researchers might consider positioning the ID so that question responses on the other side of the page will not be lost if IDs are removed (Dillman et al., 2014).

Conclusion

Survey IDs are a valuable tool for organizing survey administration, tracking progression, and managing data. Using survey IDs in Extension research can strengthen data quality, improve participant experiences, and reduce administrative challenges. Obstacles to using survey IDs can be overcome through the use of simple techniques that foster respondents' trust in researchers and increase data validity.

References

Baatard, G. (2012). A technical guide to effective and accessible web surveys. The Electronic Journal of Business Research Methods, 10(2), 101–109.

Couper, M. P., Traugott, M. W., & Lamias, M. J. (2001). Web survey design and administration. Public Opinion Quarterly, 65(2), 230–253.

Cui, W. W. (2003). Reducing error in mail surveys. Practical Assessment, Research & Evaluation, 8(18), 1–7.

Dillman, D. A., Smyth, J. D., & Christian, L. M. (2014). Internet, phone, mail, and mixed-mode surveys: The tailored design method. Hoboken, NJ: John Wiley & Sons.

Gamon, J. A. (1991). The Delphi—An evaluation tool. Journal of Extension, 29(4), Article 4TOT5. Available at: https://www.joe.org/joe/1991winter/tt5.php