Yale Program on Climate Change Communication

Survey Methods

The samples for each of the three surveys was drawn from the Ipsos (formerly GfK) 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 (non­institutional) 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 given Internet access so they may participate. These samples therefore each constitute 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 U.S. Census Bureau norms. The weights are applied when calculating all values in this report.

This report uses subsets of these samples: Americans who regularly watch, listen to, or read content from each news source. To define and create these subgroups of “the regular audience” of each news source, we asked survey respondents “How often do you watch, listen to, or read content from the following?” with individual items labeled “CNN,” “MSNBC,” “National Public Radio (NPR),” “the Weather Channel,” “the Fox News Channel,” and “the national nightly network news on CBS, ABC, or NBC.”  Response options were “Never,” “Every few weeks,” “Once a week,” “Every few days,” “Almost every day,” “Several times a day,” and “Many times a day.” For the purposes of this report, we define the regular audience of each news source as those who responded “Every few days” or more often, while excluding those who responded “Once a week” or less often. The demographics of each news source’s audience are reported on pages 20-22.

Most figures in this report display the findings from survey items that were included in all three surveys. For these items, the average margin of error is +/- 4 percentage points at the 95% confidence level. However, a few figures show findings from items that were included in only one or two of the three surveys. These smaller subsample sizes result in a greater margin of error:  an average of +/- 5 points when an item was included in two surveys and an average of +/- 7 points when an item was included in just one survey. Each figure caption notes which survey(s) contribute data to the figure. The sizes of news source audiences for each survey (and in total) are reported in Table 1.

For tabulation purposes, percentage points in each figure are rounded to the nearest whole number. As a result (and due to occasional non-response), some figures’ percentages may total higher or lower than exactly 100%. Summed response categories (e.g., “strongly agree” + “somewhat agree”) are rounded after sums are calculated (e.g., 25.3% + 25.3% = 50.6% which, after rounding, would appear in the report as 25% + 25% = 51%).

Table 1. Audience size, by news outlet