Appendix II: Survey Method

The data in this report are based on a nationally representative survey of 1,033 American adults, aged 18 and older. The survey was conducted October 20 – 26, 2023. All questionnaires were self-administered by respondents in a web-based environment. The median completion time for the survey was 22 minutes.

The sample was drawn from the Ipsos KnowledgePanel®, an online panel of members drawn using probability sampling methods. Prospective members are recruited using a combination of random digit dial and address-based sampling techniques that cover virtually all (non-institutional) resident phone numbers and addresses in the United States. Those contacted who would choose to join the panel but do not have access to the Internet are loaned computers and provided Internet access so they may participate.

The sample therefore includes a representative cross-section of American adults – irrespective of whether they have Internet access, use only a cell phone, etc. Key demographic variables were weighted, post survey, to match US Census Bureau norms.

From November 2008 to December 2018, no KnowledgePanel® member participated in more than one Climate Change in the American Mind (CCAM) survey. Beginning with the April 2019 survey, panel members who have participated in CCAM surveys in the past, excluding the most recent two surveys, may be randomly selected for participation. In the current survey, 276 respondents participated in a previous CCAM survey.

The survey instrument was designed by Anthony Leiserowitz, Seth Rosenthal, Jennifer Carman, Marija Verner, Matthew Ballew, Sanguk Lee, Matthew Goldberg, and Jennifer Marlon of Yale University, and Edward Maibach, John Kotcher, Teresa Myers, Nicholas Badullovich, and Kathryn Thier of George Mason University. The figures and tables were constructed by Emily Goddard of Yale University.

 

Sample details and margins of error

All samples are subject to some degree of sampling error – that is, statistical results obtained from a sample can be expected to differ somewhat from results that would be obtained if every member of the target population were interviewed. Average margins of error for each wave of CCAM, at the 95% confidence level, are plus or minus 3 percentage points except where noted.

  • October 2023: Fielded October 20 – 26 (n = 1,033)
  • April 2023: Fielded April 18 – May 1 (n = 1,011)
  • December 2022: Fielded December 2 – 12 (n = 1,085)
  • April 2022: Fielded April 13 – May 2 (n = 1,018)
  • September 2021: Fielded September 10 – 20 (n = 1,006)
  • March 2021: Fielded March 18 – 29 (n = 1,037)
  • December 2020: Fielded December 3 – 17 (n = 1,036)
  • Apri 2020: Fielded April 8 – 17 (n = 1,029)
  • November 2019: Fielded November 8 – 20 (n = 1,303)
  • April 2019: Fielded March 29 – April 8 (n = 1,291)
  • December 2018: Fielded November 28 – December 11 (n = 1,114)
  • March 2018: Fielded March 7 – 24 (n = 1,278)
  • October 2017: Fielded October 20 – November 1 (n = 1,304)
  • May 2017: Fielded May 18 – June 6 (n = 1,266)
  • November 2016: Fielded November 18 – December 1 (n = 1,226)
  • March 2016: Fielded March 18 – 31 (n = 1,204)
  • October 2015: Fielded September 30 – October 19 (n = 1,330)
  • March 2015: Fielded February 27 – March 10 (n = 1,263)
  • October 2014: Fielded October 17 – 28 (n = 1,275)
  • April 2014: Fielded April 15 – 22 (n = 1,013)
  • November 2013: Fielded November 23 – December 9 (n = 830)
  • April 2013: Fielded April 10 – 15 (n = 1,045)
  • September 2012: Fielded August 31 – September 12 (n = 1,061)
  • March 2012: Fielded March 12 – 30 (n = 1,008)
  • November 2011: Fielded October 20 – November 16 (n = 1,000)
  • May 2011: Fielded April 23 – May 12 (n = 1,010)
  • June 2010: Fielded May 14 – June 1 (n = 1,024)
  • January 2010: Fielded December 24, 2009 – January 3, 2010 (n = 1,001).
  • November 2008: Fielded October 7 – November 12 (n = 2,164).
    • Data were collected over two periods: from October 7 – 20 and from October 24 – November 12. Margin of error plus or minus 2 percentage points.

 

Rounding error and tabulation

In data tables, bases specified are unweighted, while percentages are weighted to match national population parameters.

For tabulation purposes, percentage points are rounded to the nearest whole number. As a result, percentages in a given figure or table may total slightly higher or lower than 100%. Summed response categories (e.g., “strongly agree” + “somewhat agree”) are rounded after sums are calculated. For example, in some cases, the sum of 25% + 25% might be reported as 51% (e.g., 25.3% + 25.3% = 50.6%, which, after rounding, would be reported as 25% + 25% = 51%).