The Faith Communities Today (FACT) surveys were coordinated by the Hartford Institute for Religion Research at the Hartford Seminary. Forty-one faith groups and denominations participated in the project. Each group designed a questionnaire in order to collect data that could be used to compare the local churches, synagogues and mosques within or between the participating groups. This dataset represents the combined results from the FACT surveys.
- Data File
- Cases: 14,301
Weight Variable: WEIGHT
- Data Collection
- Date Collected: 2000
- Funded By
- The survey was funded by The Lilly Endowment, Inc., with matching funding from the participant groups.
- Collection Procedures
- FACT2000 is the largest survey of congregations ever undertaken in the United States. It was a cooperative effort among twenty-six agencies and organizations representing forty-one denominations and faith groups – from Southern Baptist to Bahai, Methodist to Muslim to Mormon, Assemblies of God to Unitarian Universalist, Orthodox, Roman Catholic, Jew, and all the usual oldline Protestant players. These groups represent an estimated 90% of all U.S. congregations. The groups worked together to develop a common, key-informant questionnaire containing 190 items. Groups then adapted wordings to their respective traditions and conducted their own survey, typically mailed during 2000 to a stratified random sample of a group’s congregations. Return rates averaged over 50%. Independent congregations proved their independence with the lowest rate of return and the Church of Jesus Christ of Latter-day Saints demonstrated one of the virtues of hierarchy with a 98% return rate. Data from the total of 14,301 returns from the various group samples were returned to the Hartford Institute for Religion Research coordinating office for aggregation. Zip-code-level census data was added to each congregational case. More detailed information about participants and methodology, as well as an electronic copy of the original FACT report and updates about the more recent surveys and resources of the Cooperative Congregational Studies Partnership (CCSP) can be found on the FACT website: http://www.fact.hartsem.edu
- Sampling Procedures
- The project’s sampling guidelines encouraged each denomination/group to use a stratified random sample of congregations with a sampling error of, at a minimum, plus or minus four percentage points at the 95% confidence level. More specifically, groups for whom it was possible were encouraged to stratify by region and size of congregation. Groups that did not have population data on size of congregation were encouraged, nevertheless, stratify by region. These guidelines were followed with only a few exceptions. All but one of the exceptions drew straight or systematic random samples from their population lists. The sample for the Historic Black Denominations was drawn by the Gallup Organization, and the specific methodology is not known. The Gallup organization was used because the Interdenominational Theological Center (ITC) which represented the Historic Black Denominations in the project felt a telephone survey conducted by a nationally prominent organization would maximize responses from black congregations. ITC and Gallup worked collaboratively to develop as inclusive a list as possible of telephone numbers for black congregations. Gallup sampled randomly from this list until it had reached it targeted number of returns.
Known or estimated total number of congregations in a denomination or faith group, sample sizes and response rates can be found in the “FACT2000ArchiveWeights” file.
From the perspective of the total number of congregations represented by all the denominations/groups participating in FACT2000, our approach to sampling amounts to stratifying by denomination/group and then sampling disproportionate to denomination/group strata size. This is unproblematic for the analysis of any given denomination/group’s data. However, for the all-group-aggregate dataset it requires the calculation and inclusion of weights to adjust for the otherwise disproportionate-to-denomination/group-strata size. Such weighting procedures for disproportionate-to-size strata are common statistical practice, but they do require conscious attention. It is imperative to note that the SPSS dataset default is set to weight the data to the actual proportion of a denomination or faith group’s congregations in the United States. We have every reason to believe that the weighted dataset represents a reasonable national sample of congregations in the United States of the forty-one denominations/faith groups involved in CCSP. We estimate that these groups represent just over 90% of congregational members in the U.S. The denomination/group weights used and their calculation are contained in the “FACT2000ArchiveWeights.”
- Principal Investigators
- Hartford Institute for Religion Research
Cooperative Congregational Studies Partnership
- Related Publications
- More detailed information about participants and methodology, as well as an electronic copy of the original FACT report and updates about the more recent surveys and resources of the Cooperative Congregational Studies Partnership (CCSP) can be found on the FACT website: http://www.fact.hartsem.edu
- The FACT2000 questionnaire was collaboratively designed by research and program staff representatives from the twenty-six agencies and organizations that ultimately used it to survey their constituent congregations. Four general principles framed the process. One was that a broad coverage of areas of congregational life was preferable to a narrowly focused survey. A second was that any specific item included in the core questions be directly identifiable as of high interest to one or more of the project’s target user constituencies, which included congregational leaders, denominational/group resourcers of congregations, the general public through the media, and academic researchers. A third was that questions in the common core questionnaire would focus on items that were applicable to the vast majority of participating groups to maximize the comparative perspective of the project. The fourth was that length of the questionnaire be affordable and not overly detrimental to a reasonable return rate. Once the decision was made to use a key-informant approach: (a) it was refined to a goal of ten pages and/or about twenty minutes, and (b) seven areas of congregational life to cover in the core questions were articulated. These seven areas included:
1. Spiritual and organizational vitality;
2. The variety and style of worship — the foundational act of religious gathering;
3. The variety of congregational activities/programs which nurture faith or provide opportunities for the expression of faith;
4. Levels of participation and the characteristics of participants;
5. Strategies congregations use to reach new members and raise financial resources;
6. Characteristics of clergy and lay leadership;
7. How congregations relate to other congregations, to denominational structures and to other institutions in their communities; and,
8. The widely different ways that congregations support and strengthen the social and material well-being of their communities.
The result was the 190 item “common core” questionnaire contained in the “FACT2000ArchiveQuestionnaire” file. A one-page outline of the 190 items can be found in the “Fact2000QuestionOutline” file.
It is critical to reiterate that with the common core questionnaire as the baseline, each group was encouraged to adapt the questionnaire to the language and traditions of its constituent congregations. Three guidelines directed the adaptation process:
1. All participating groups would use all the core items, except those items for which a group could articulate a compelling reason to omit;
2. Each group needed to adapt the wording of core items tits respective traditions (e.g., imam, minister, president, priest, rabbi); and
3. Each group was free to supplement the core with questions of its own unique interests.
In most cases “adaptation” involved changing an occasional word or phrase -- e.g., pastor, rabbi, imam, priest; congregation, church, synagogue, mosque; etc. However, in a few instances a common question was deleted by a particular group, or was so changed that it was incompatible. Copies of the adapted questionnaires used in most of the surveys can be viewed in the ‘Partner” section of the FACT website --http://fact.hartsem.edu/partners/index.html.
While it appears fair to say that the adaptation guidelines were followed in the vast majority of instances, it also appears fair to say that there were some exceptions. It is also true that the circumstances of a few participating groups dictated a relatively radical adaptation process, this arguably being most true for the Historic Black denominations, Muslim, Baha´is and Roman Catholic. One consequence is that the data from a particular denomination/group may be missing for some variables. It is also the case that the data for a few questions in the common core, primarily the specific financial questions, appeared so uneven that the entire question was deleted from the dataset. Additionally, it is important to note that a congregation’s specific denomination or faith group is NOT contained in the archive dataset, only a five category, denominational family variable (see the “FACT2000ArchiveWeights” file for how FACT2000 denominations and faith groups were categorized). Specific denomination/group is omitted to help protect the anonymity of individual congregations.
Each denomination or faith tradition was free to choose the best key informant to complete the questionnaire. In all cases this was the senior pastor, rabbi, imam, priest, etc., or in their absence, the senior “lay” leader (e.g., president of a Muslim Center or Mormon Ward).
- Zip Code Level Census Data
- There is a near consensus among scholars and practitioners that the life and mission of congregations cannot be adequately understood apart from the social context within which a congregation is embedded. To provide both a reminder of the importance of the social context and concrete data on the most geographically immediate social context of congregations, U.S. Census data for the zip-code in which a congregation is located has been merged each congregation’s questionnaire data in the archived dataset. The merged zip-code-level census data was purchased from Applied Geographic Solutions, Inc (AGS), and includes a few items from 1980, a more comprehensive list of items for 1990 and 2000, and sets of 2005 and 2010 projections. Because the 2000 U.S. census data was not available at the time the aggregated FACT2000 dataset was created, the 2000 data is also a projection. A list of merged census variables is contained in the “Fact2000ZipAggVars” file. Files documenting AGS methodology and terminology are also available.
While one will find zip code data and a variety of census geography and locational type variables in the dataset, one will not find a congregation’s actual zip code. As was the case for a congregation’s specific denomination or faith group, a congregation’s specific zip code has been omitted from the dataset to help protect the anonymity of individual congregations.