About the survey

Methodology

This survey was conducted online among a sample of 1,442 self-identified Catholic adults who are part of the Knowledge Networks’ KnowledgePanel. The KnowledgePanel is a nationally representative probability sample of the U.S. adult population. Panel respondents who do not have Internet access are provided with Internet service and free laptop computers by Knowledge Networks, which ensures that panel respondents are representative of the national population and are not limited only to those who already use the Internet. Additional details about the KnowledgePanel are available on the Knowledge Networks website at www.knowledgenetworks.com/knpanel/KNPanel-Design-Summary.html.

Interviewing took place April 25-May 2, 2011. The survey included extra interviews (oversamples) with both Hispanic Catholics and young Catholics (those under age 32). At the direction of the research team, no interviews were conducted with Hispanic Catholics over the age of 70, who comprise roughly 1 percent of the U.S. Catholic population. The survey results are thus representative of the views of approximately 99 percent of the U.S. adult Catholic population.

The data were adjusted (i.e., weighted) to correct for the extra interviews with Hispanics and younger Catholics, so that these groups are represented in the final survey results in proportion to their actual share of the Catholic population. Additionally, the demographics of the sample (gender, age, race and ethnicity, primary language, educational attainment, geographic region, metropolitan area and Internet access) were weighted to ensure that the demographic characteristics of the sample match the actual demographics of the Catholic population.

Interviews were conducted in both English and Spanish. Conducting interviews in Spanish is particularly important when seeking to represent the views of populations that include large numbers of Hispanics, since many Hispanics are recent immigrants who are unable to complete a survey in English. More than half of the Hispanic Catholics surveyed (56 percent) completed the interview in Spanish.

Of all of the Knowledge Networks panelists who were invited to take the survey, 49 percent completed the interview. The overall response rate for the survey, which takes into account the rate at which those who were recruited to join the Knowledge Networks panel agreed to do so, is 5 percent. Results based on the full sample of 1,442 Catholics have a margin of sampling error of plus or minus 3.5 percentage points (at the 95 percent level of confidence); the margins of sampling error for subgroups of Catholics (such as Hispanic Catholics or millennial Catholics) are larger. Readers should bear in mind that in addition to sampling error, other factors such as question wording and the ordering of the questionnaire can introduce error or bias into public opinion polls. In addition, the different mode of administration of the current survey (self-administered, online survey) compared with the previous surveys (interviewer-administered, telephone surveys) means that care must be exercised when interpreting change over time. Some respondents may give different responses to a question when an interviewer is present than when no interviewer is present, especially when a particular type of response might be seen as socially desirable. As a result, we have interpreted trends cautiously in our analyses, and relied on patterns in change over time observed in the first four surveys to help us interpret the meaningfulness of apparent changes in the current data.

All percentages used in the essays in this section are rounded to the nearest whole number. This may occasionally lead to a 1 percentage point difference between two percentages given in the body of the essay, and the sum of the two as found in the table or other figure.