Uploaded November 2015 | Updated September 2026, 2 hours ago
Our lives are full of results from surveys—health information, demographic changes, popularity of products, who is likely to be nominated to run for President. How do you know when you can trust the results of survey data? What do you need to look for?
This talk will cover many ways that survey results can be tainted. The sampling of survey respondents might be biased. The question wording might produce incorrect data. The survey may be conducted using inappropriate methods that produce unreliable data. Examples will be drawn from current news articles and web sites that use surveys including election forecasts and government survey results. We will also talk about fund raising under the guise of surveys (frugging) and why elections predictions can fail.
This talk will give you the tools you need to judge the findings of surveys. You will be able to understand why the lack of a discussion of survey methods is often a way to hide unreliable results. You will also understand more clearly why surveys are not always the right method for conducting research.
Fran Featherston is retired from the U.S. federal government where she worked 27 years as a researcher at the National Science Foundation (10 years) and the Governmental Accountability Office (17 years). Her areas of expertise are survey research, research design, and designing methods that are user friendly. Kristin Stettler has worked as a federal researcher for 22 years. She currently works at the U.S. Census Bureau, where she works with Census staff to design surveys to collect the most accurate data. Kristin and Fran have taught seminars on survey design at two international conferences. They will bring their expertise to discuss what you need to know to assess the validity of survey data.
Views expressed are those of the speakers and do not necessarily reflect the views of the National Capital Area Skeptics.
Our lives are full of results from surveys—health information, demographic changes, popularity of products, who is likely to be nominated to run for President. How do you know when you can trust the results of survey data? What do you need to look for?
This talk will cover many ways that survey results can be tainted. The sampling of survey respondents might be biased. The question wording might produce incorrect data. The survey may be conducted using inappropriate methods that produce unreliable data. Examples will be drawn from current news articles and web sites that use surveys including election forecasts and government survey results. We will also talk about fund raising under the guise of surveys (frugging) and why elections predictions can fail.
This talk will give you the tools you need to judge the findings of surveys. You will be able to understand why the lack of a discussion of survey methods is often a way to hide unreliable results. You will also understand more clearly why surveys are not always the right method for conducting research.
Fran Featherston is retired from the U.S. federal government where she worked 27 years as a researcher at the National Science Foundation (10 years) and the Governmental Accountability Office (17 years). Her areas of expertise are survey research, research design, and designing methods that are user friendly. Kristin Stettler has worked as a federal researcher for 22 years. She currently works at the U.S. Census Bureau, where she works with Census staff to design surveys to collect the most accurate data. Kristin and Fran have taught seminars on survey design at two international conferences. They will bring their expertise to discuss what you need to know to assess the validity of survey data.
Views expressed are those of the speakers and do not necessarily reflect the views of the National Capital Area Skeptics.
