Back to top

The Why and How of Public Opinion Research

I Introduction

Most organizations depend on the support of members, stakeholders, donors and/or the general public to accomplish their mission. It is therefore crucial that organizations have a complete understanding of the opinions, needs, concerns and desires of these key groups. Often, organizations mistakenly believe they know what their clients and stakeholders want or need. It is critical to step out of the organization's head space and check in with these groups. Public opinion research is one way to do this. It can be a powerful planning and evaluation tool in many situations. Some of these situations are outlined below, along with an overview of how to conduct public opinion research.

II Why Conduct Public Opinion Research?

Public opinion research can help organizations understand their clients/audiences. This audience information can be used for a wide variety of planning and evaluation tasks. For example

* Establishing an understanding of how various groups (i.e., supporters, potential supporters and opponents) view an organization can help with efforts to clarify the mission, vision and activities.

* Increased knowledge about how clients view certain issues can help with the development of advocacy messages and strategies. For example, a recent poll regarding the public education system revealed that parents support reducing class sizes (also supported by teachers) and more emphasis on the basics (also supported by the government). This information provided both teachers and the government with clues about which of their policy change suggestions would garner the most support.

* Identifying the difference between one-time and regular donors can help with the creation of targeted fundraising strategies and ultimately increase revenue.

* Exploring client definitions of "success" can help with strategic planning, program planning and evaluation efforts.

* Understanding client needs and wants can help improve and create messages, activities and ways of work (e.g., communication methods).

* Releasing public opinion research to the media can be a cost effective method for attracting media and public attention. It should be noted, however, that the media may have relatively high expectations in terms of methodology and may not report on research that uses a small sample size.

~ * ~ * ~ * ~ * ~ * ~ * ~ * ~ * ~ * ~ * ~ *

III How is Public Opinion Data Collected?

There are many ways to carry out public opinion research, including written surveys, telephone surveys, key informant interviews, focus groups and participant consultations (in which both the researchers and the respondents discuss a problem or issue). Both quantitative and qualitative methods are important. Quantitative data can provide clear numbers but qualitative data can provide the story behind the numbers. Qualitative methods, such as focus groups or open-ended written questions, can also be used prior to quantitative methods to identify which questions to ask using quantitative methods (e.g., multiple choice surveys).

Sample size and sampling method are as important as the data collection tool itself. Sample size should be based on the level of accuracy required. Larger sample sizes increase accuracy and allow analysis of subgroups within a sample (for example, to draw separate conclusions about men and women a sufficient number of each are required). Budget is also an important factor because larger samples cost more!

Truly random sampling methods combined with a large sample size can allow generalization to the entire target population. However, random sampling can be more expensive and complicated to achieve than convenience samples of "whoever shows up" or "people you know."

~ * ~ * ~ * ~ * ~ * ~ * ~ * ~ * ~ * ~ * ~ *

IV Other Considerations

Collecting high quality public opinion research information can be challenging. Here are a few tips to consider before planning a public opinion research project.

1. Be optimistic. Don't start by assuming you are unable to afford it.

2. Join forces. Form a partnership with other related organizations to conduct data collection.

3. Take your time. Spend time getting to know the different data collection options. Carefully think through the questions you want to ask. Make sure the research method and your questions are appropriate for gathering the information you need.

4. Focus. Set specific, clear goals for your research project.

5. Don't do everything at once. Spread your research out over a number of years (a multi-phased approach). Perhaps conduct focus groups one year and a survey the next.

6. Look for opportunities. Adding one question to an Omnibus poll costs a relatively small amount.

7. Set expectations. If you are working with a consultant, expect practical recommendations not just a stack of data tables.

8. Test your tools. It is always a good idea to carry out at least a small focus group to pretest your survey questions and general research design. This will help identify any ambiguous questions and reveal whether you are in fact getting at the information you wanted, etc. When you run a survey, always test it in the field for a few hours and make sure your interviewers report-back on any problems of logic, understandability, format and so on.

9. Balance accuracy with what is realistic. The recent surge of media polls has contributed to the belief that a margin of error +/- 5 percent, with 95 percent confidence (i.e., 19 times out of 20) is a standard. This small margin of error and high confidence may be ideal, but it is not always necessary. This level of precision will increase the cost of your research because a large sample is required. Think about what level of confidence is acceptable for your situation. If the research is for strategic planning, for example, the level of confidence can be lower than it would be in an academic setting. How many management decisions require 95% confidence? If they did, not much would ever be decided by managers!