The FACT 2005 survey questionnaire was designed by the Research Taskforce of The Cooperative Congregational Studies Partnership. A copy of the questionnaire is available here and should be consulted for exact question and response category wording.
The Center for Social Research at Calvin College conducted the survey. The questionnaire was mailed to a random sample of 3,000 congregations, and the accompanying cover letter also included the option of completing the questionnaire online. The sample was originally generated by American Church Lists, then reviewed and cleaned by CCSP denominations and faith groups. Random replacements for non-responding congregations were drawn from an American Church Lists shadow sample and from denominational yearbook samples. 884 usable questionnaires were received.
To enhance national representation, responses were weighted to the population parameters for region and faith family presented by Hadaway and Marler [C. Kirk Hadaway and Penny Long Marler, How Many Americans Attend Worship Each Week? An Alternative Approach to Measure, Journal for the Scientific Study of Religion (2005) 44(3): 307-322, Table 2], and for size of congregation and rural/city/suburban location found in the FACT 2000 national survey of 14,301 congregations.
Sampling error for such a survey can only be estimated. We estimate it to be +/- 4% at the 95% confidence level.
In several places in the following report we present comparisons across Faith Families. These comparisons typically are among Oldline Protestants, Other Protestants and Catholic & Orthodox. A more specific definition of these families can be found below. The total of 884 usable questionnaires does include Jewish, Muslim and other non-Christian congregations, but in a sample of this typical national survey size they are too few and diverse to reliably report as a distinct group. They are included in the figures representing the total sample, but we are unable to include them in our Faith Family comparisons.
The FACT 2000 survey, in contrast, did include meaningful samples of Baha’i, Jewish and Muslim congregations, and one can find references to reports on these Faith Families on the FACT website. Similarly, one can find references to reports on the FACT 2000 survey of Historically Black Protestant denominations on the FACT website. A survey the typical size of FACT 2005 must use Faith Families because it does not include sufficient numbers of congregations from any single denomination for separate analysis. However, several CCSP member denominations conducted oversamples of their congregations to provide the base for their own reporting. These included The Church of the Nazarene, The Episcopal Church, and The Unitarian Universalist Association.
View the common core questionnaire used to collect information for the FACT 2005 study.