We’ve noted in previous posts that it is important to follow stringent guidelines when conducting a contingent valuation (CV) survey because it can be crucial to determine accurate measurements. Our research suggests that following these guidelines results in CV estimates that eliminate bias and provide a range of value impacts that are consistent with impacts determined through other methods. Two of the most notable published guidelines that we follow are those of the National Oceanic and Atmospheric Administration (NOAA) Blue Ribbon Panel on Contingent Valuation[1] and those of Diamond in her “Reference Guide on Survey Research.”[2] Both studies indicate the importance of following various guidelines for accurate survey results, and both are loosely formatted so that these guidelines can be implemented at different stages of research.

As we used these studies in our work, we began to notice that the guidelines could generally be divided into three groups to be checked at the following research stages:

  1. Determining a Survey Sample Universe
  2. Creating the Survey Instrument
  3. Collecting, Analyzing, and Reporting the Data

Determining a Survey Sample Universe

Today’s post focuses on the guidelines for determining a survey universe, one of the initial stages of research for CV surveys. In this stage, guidelines focus on who should take the survey and how to include them as part of the sample. Before conducting a survey, researchers decide on the most appropriate group of people to survey. This entire group from which the sample will be taken is referred to as the survey universe. Determining an appropriate survey sample is important because you want to make sure that the survey sample matches the study area population as closely as possible. In survey research, when the survey sample is representative of the study area population, it helps to eliminate any differences between the groups that might affect how individuals value the condition being studied. For example, if we were conducting a project in Hawaii, it would not make sense to survey folks in Nebraska. There is a good chance that folks in Nebraska do not have experience with the good being valued in Hawaii, especially if it is a good that has cultural significance. We created the figure below to help illustrate these concepts.

It is important to note that in most cases, you cannot choose the study area population itself as the survey universe because the peer-reviewed economic literature suggests that individuals who may have a stake in a given situation will have incentive to act strategically. For example, there may be a town that has a neighborhood with contaminated drinking water wells, the effects of which we want to study. Our study area population would be the affected neighborhood; our survey universe would be all neighborhoods in the town (and likely adjacent towns) that do not have a contaminated well, and the survey sample would be a portion of that survey universe. Therefore, we choose a survey universe that is similar to the study area population but that is not directly affected by the condition being measured.

One way to acquire a survey sample similar to the study area population is with demographic screening questions, though there are many others. The screening question might ask the respondent if they own or rent their home. If a respondent indicates that they rent their home, they may be terminated from completing the survey. This is because renters do not have as much experience with the home-buying process as someone who is currently a homeowner who will likely have a better understanding of how something such as contamination could impact their property’s value. Although screening questions are usually straightforward, the process of designing, writing, and presenting survey questions can affect survey results. In upcoming posts, we will discuss the guidelines for these other research stages.

This post has covered only the basics of determining a sample universe. For more information, feel free to use ourAsk an Expert page for any questions.

– Abigail Mooney and Sarah J. Kilpatrick

 

1. Arrow, K., R. Solow, P.R. Portney, E.E. Leamer, R. Radner, and H. Schuman. 1993. “Report of the NOAA Panel on Contingent Valuation.” National Oceanic and Atmospheric Administration. Federal Register 58:4601–4614.  
2. Diamond, Shari S. 2011. “Reference Guide on Survey Research.” pp. 359–424 in Reference Manual on Scientific Evidence, 3rd Edition. Federal Judicial Center. Washington, DC: National Academies Press.