Washington Redskins quarterback Kirk Cousins on the sidelines this past fall at FedEx Field in Landover Md. (John McDonnell/The Washington Post)

A new Washington Post survey explores attitudes of Native Americans toward the Washington Redskins team, marking the first national gauge of the group’s views of the issue in more than a decade. The following questions and answers address some of the major issues in surveying Native Americans and how this poll was conducted.

Q: Why is it difficult to survey Native Americans?

Surveying the Native American population is difficult because of the group’s relatively small size and the fact that many who live on reservations lack landline telephone access. A 2004 survey by the Annenberg Public Policy Center, which found 90 percent of Native Americans were not offended by the Redskins name, has been criticized for potentially underrepresenting Indians who live on reservations and are less likely to have landline phones, for not measuring levels of tribal membership and for only asking a single question about attitudes on the issue.

The Washington Post survey was designed to overcome challenges in surveying Native Americans by reaching a large portion of the sample through cellular phones. Roughly 95 percent of Native American adults have landline or cellular phone access in their households, and over half are cellphone-only, according to The Post’s analysis of the National Health Interview Survey. Nearly 6 in 10 Native Americans in The Post survey were interviewed on a cellular phone.

The Post survey also asked questions about tribal membership and several questions to capture attitudes toward the team’s name and the broader use of Native American imagery in sports. In addition to standardized survey interviews, The Post conducted more than two dozen follow-up interviews with survey respondents who agreed to speak with reporters during the initial survey.

Q: How were Native Americans chosen for the survey?

The survey was conducted in conjunction with weekly national surveys of U.S. adults reached through a random sample of cellular and landline telephones conducted by SSRS of Media, Pa. Toward the end of a survey on a range of topics, respondents who identified their race as Native American were asked a series of questions on views of the Redskins team name and Native American imagery in sports. While Native Americans account for about 2 percent of the total population, the survey’s extended field period from December to April reached 504 Native American respondents. The survey interviewed Native Americans living in all 50 states, including Alaska and Hawaii, and respondents both living on or near reservations and those who do not.

Q: How did the survey define Native Americans?

All survey respondents identified themselves as Native Americans, American Indians or Alaska Natives when asked, “Do you consider yourself white, black or African American, Asian, Native American, Pacific Islander, mixed race or some other race?” While many multi-racial Americans have partial Native American ancestry, the survey focused only on those who first identified themselves as Native Americans. This definition most closely compares with the Census Bureau’s categorization American Indian Alone, rather than American Indians in combination with another race.

Q: How did the survey account for tribal membership?

Respondents were asked whether they are currently enrolled as a member with a Native American tribe and to which specific tribe they belong. Tribal members represented 36 percent of interviews conducted and accounted for 44 percent of the final weighted sample, which matches the Census Bureau’s data on demographic and geographic characteristics. Cherokee and Navajo members accounted for the largest share of this group, though most reported enrollment in smaller tribes. Survey results are reported both for tribal members and non-tribal members, as well as for other demographic groups.

Q: How did the survey reach Native Americans living on reservations?

The survey used respondents’ Zip codes and the 2010 Census to identify whether their neighborhood included a federally recognized reservation or tribal land. Roughly 1 in 10 interviews were conducted among Native Americans who lived in Zip codes where at least 75 percent of land is on a reservation or tribal area, and another 10 percent live in Zip codes with at least some portion of tribal land. Final survey results were weighted to match census data indicating 20 percent of single-race Native Americans live in Zip codes with at least 75 percent reservation land, and 17 percent live in Zip codes containing at least some reservation land.

Q: How did the survey ask whether the Redskins name is offensive?

Questions must be clear to all respondents and avoid leading language to produce accurate results. The survey’s first question about the Redskins team name asked, “The professional football team in Washington calls itself the Washington Redskins. As a Native American, do you find that name offensive, or doesn’t it bother you?” This same question was asked in a 2004 Annenberg Center survey, and some have suggested respondents could misunderstand the phrase “or doesn’t it bother you?” The Post asked follow-up questions of the first 43 respondents to verify they understood the questions and found no respondents changed their answer due to confusion about the wording (see methodology for more detail on this test). The results from this question are also broadly in line with other questions in the survey asking the general offensiveness of the word “Redskin.” Exact question wording, order and percentage results can be found at wapo.st/pollarchive.

Q: Can a sample of 500 Native Americans be projected to the entire population?

Surveys of this size can produce accurate estimates of any size population by employing “probability sampling,” where every member of the population has a known and equal chance of being selected. Such surveys are used to measure consumer confidence, the unemployment rate, political attitudes and health. The Post’s survey of 504 Native Americans has a margin of sampling error of 5.5 percentage points, meaning that if the same survey were repeated 100 times, in 95 cases the results would not be expected to differ by more than 5.5 percentage points. Surveys can have other sources of error and variation which are more difficult to quantify, including measurement error and non-response error and the fact that about 5 percent of Native Americans are unreachable by either cellular or landline phone. Statistical weighting to Census Bureau benchmarks helps correct for some of these issues.

Q: Is the survey demographically representative of the Native American population?

Yes. Survey results were statistically weighted to match U.S. Census Bureau estimates of the demographic and regional population characteristics of single-race American Indian/Alaska Native adults. The final sample matches population estimates for gender and age groups, Hispanic ethnicity, educational attainment, regional makeup (Northeast, North-Central, South, Mountain and Pacific), and proximity to reservation and tribal lands.

Full question wording and additional technical information on the survey methodology is available here.