Appendix B - Methodology

Methodology

The Generosity Gap research consists of data and analysis based on three phases of study. First, in February 2016, Barna conducted 81 qualitative, or open-ended, surveys with U.S. adults ages 18 to 69. These participants were recruited through a national consumer panel that is representative by age, gender, ethnicity, region and socioeconomic level. However, these results are not intended to be representative of, or projected to, the adult population. Also as part of the qualitative phase, 21 Protestant pastors completed open-ended surveys in March 2016 through Barna’s PastorPanel. These pastors are broadly distributed across denominations, church size and region, but again, not intended to be nationally representative.

Second, quantitative research was conducted with 1,556 self-identified Christians who have at- tended church at some point in the past year and who agree, or are neutral, that their faith is very important in their life. These adults represent a range of denominations, including Catholic, mainline and non-mainline Protestant. Participants for this “interested Christians” sample were recruited from a national consumer panel in July and September 2016. The total number of respondents included an oversample of 747 Millennials to enable more robust analysis of this generation, but the total sample was then weighted to be representative of the U.S. adult population by age, gender, ethnicity, region and socioeconomic grade. The sampling error on this survey is plus or minus 2.3 percentage points at the 95-percent confidence level.

Third, researchers designed a quantitative survey for U.S. Protestant pastors; 606 completed the survey in June 2016. Participants were recruited through Barna’s PastorPanel, publicly available lists and email invitations. Data were minimally weighted to be nationally representative of Protestant churches by denomination, church size and region. The sampling error on the pastor survey is plus or minus 3.9 percentage points at the 95-percent confidence level.