When it comes to answering surveys, most folks don't mind telling you what kind of car they drive or what kind of ice cream they eat. But if you start asking about their religious preference, they clam up. Heck, even the U.S. Census doesn't try to gather religious data!
So how do you get people to not only divulge their choice of religion but talk about it in-depth? And how do you cost effectively locate people whose choice of religion puts them in a distinct minority (2% of the population)?
In 1989, these were just two of the questions facing the Council of Jewish Federations (CJF) and the City University of New York when the two organizations were considering undertaking the National Jewish Population Study (NJPS), a major survey of the views, practices and lifestyles of the U.S. Jewish population.
The CJF is an umbrella organization for the almost 200 Jewish federations serving more than 800 cities across the U.S. and Canada. The federations and many other Jewish organizations provide a large number of social services, including educational, health and recreation programs.
National level
While many of the individual local federations have performed studies of the users and potential users of these services, there really wasn't any data available on a national level, says Barry Kosmin, director of research for the Council of Jewish Federations.
"There are a lot of local studies, but it's like looking at the win-loss situation for your favorite sports team. It doesn't tell you anything unless you know how the whole league is doing. A national survey is a baseline against which everybody can compare themselves."
Kosmin, who is also director of the Berman Institute at the graduate school of the City University of New York, served as director of the National Jewish Population Study
"The Jewish population operates a sophisticated voluntary sector which needs to know something about its constituents. Beyond the synagogues and religious education, the Jewish community is a major fundraising organization. It operates hospitals and family services, Jewish community services, vocational services, youth programs. It has a whole social agenda," Kosmin says.
To find Jewish respondents to participate in the study. the CJF used EXCEL, a telephone omnibus study conducted by AUS Consultants-ICR Survey Research Group in Media, Pennsylvania. EXCEL reaches 2000 respondents each week through twice weekly interview waves. Each study typically covers a variety of topics from current events to questions on products and services. Some of the resulting data is shared, but subscribers can include proprietary questions to collect data for their own private use.
Very efficient
A national omnibus study can be an efficient way to screen for hard-to-find respondents because while the study may have several thousand respondents, the cost of reaching those respondents is spread out over several subscribers.
As a subscriber, you can include questions designed to ferret out your target respondent. For example, Mannington Resilient Floors and Holiday Corp. have used omnibus mail panels to screen for specific consumer segments.
When Mannington was considering designing a new flooring product, it wanted to survey owners of vinyl floors. And when Holiday Corp. wanted to test its Homewood Suites lodging concept, it sought the opinions of business travelers. Both consumer groups were relatively small, but through questions in an omnibus study, the two companies identified a large sample of people in the desired consumer segments, and they were able to contact them later for a more in-depth survey.
"For the Jewish population study," Kosmin says, "the question was, how do you screen large numbers of people to get a representative sample of a rare population? If you went through the streets of New York City you could find thousands of respondents, but if you want to find Jews in Montana, that's different.
"The beauty of the omnibus is the national perspective that you get. We wanted to speak to a rare population. We were trying to be as objective as possible. We weren't going to use our own lists to find respondents because people could turn around and say that we had hand chosen the people. The selection had to be random. Mail surveys and door-to-door weren't a practical proposition. This method seemed to be the best, at least in terms of resources expended--namely time and money."
Nationwide study
During the EXCEL interviewing respondents were asked about their religious preference as part of a series of demographic questions. This information not only served as a screening device for the Jewish population study, it was also the basis of the landmark National Survey of Religious Identification (NSRI), a nationwide study of religion affiliation. (The NSRI was conducted under the auspices of the Graduate School and University Center of the City University of New York.)
The NSRI results were based on over 113,000 interviews conducted from April 1989 to April 1990, says John De Wolf, account executive, AUS Consultants-ICR Survey Research Group. "EXCEL provided a good vehicle--since we were going to ask about religion anyway--to do a study on religion in general in the population which has never been available before on this scale."
Throughout the EXCEL interviews, only 2.3 percent of the respondents refused to answer the question about religion. De Wolf says that this low refusal rate can probably be attributed to the fact that the interviewers had gained the trust of the respondents by the time the religion questions were asked late in the survey.
"The EXCEL survey usually starts with some very topical questions pertaining to current news or political events. That has a tendency to legitimize the study. Then you get into some other topics such as videotape rentals or toothpaste usage, so by the time we get to the demographics we've talked about several different subject matters and the respondents know we're not trying to sell them anything.
"That's one of the main problems with a subject as personal as religion. If you wanted to do this on a custom basis and call people to ask what their religion is, you would lose most of them on refusals."
Re-contacted
After suitable respondents for the Jewish population study were identified through the omnibus study, they were re-contacted to determine their interest in participating in 30 minute telephone interviews that would form the basis for the NJPS.
Those interviews were conducted with 2,500 households between May and June 1990. The survey included a variety of questions on attitudes, beliefs and practices, including observance of holidays, the level of Jewish and civic attachment, and feelings about Israel, in addition to demographic and economic information on income, household composition, marital status.
The survey contained specific questions about fundraising, what their attitudes were towards fundraising efforts, how they wanted to be solicited, what they saw as priorities, etc. "If you cross-tabulate that with the demographic information it gives fund-raisers a tremendous amount of insight into how to operate in the marketplace," Kosmin says.
That information has two uses: It guides fundraising efforts by identifying those who are most interested in giving, and it aids allocation of funds by identifying potential need for services. "There is the need to forecast, for instance, day care needs, so there are questions in the survey that enabled us to find out how many women are going to have children in the next three years. So a Jewish community center or a synagogue nursery, for example, can measure whether there's an increased need for their services."
Friendly context
Asking the religious questions as a part of the multi-faceted omnibus study helped the response rate by placing the potentially intrusive questions in a friendly context, Kosmin says.
"When we asked people about their religion, they seemed reasonably happy to tell us. They didn't regard it as a terribly threatening question. The omnibus survey is more helpful than a direct survey, in which an interviewer might call the respondent and ask 'what is your religion?' and then proceed with 15 questions about those issues. That can make people much more hesitant than if the questions are asked as part of a multi-purpose survey. It's just less threatening."
Same climate
In addition, it was important to complete the interviews for the NJPS in a short period of time, to make sure the respondents were answering the questions in roughly the same political and economic climate. "If you're interested in things like the welfare situation, you don't want to drag the interviewing out over two years because the economic situation might change. We wanted to look at attitudes towards Israel and people's level of emotional attachment, and if we'd done some of the interviews before the Gulf war and some after, obviously there would have been a problem."
Strong support
The data is being analyzed by a number of sources, and several reports taking a more in-depth look at specific sections of the data are scheduled for release.
The survey has been met enthusiastically and there is strong support for doing the survey again in five to ten years, Kosmin says. Like any product or service provider, the Jewish voluntary sector and the individual federations need to keep in touch with their "market."
"The fundraising of the Jewish community in the United States and its organizations plus the services that it provides to people, you're talking about around $3 billion. That's why you can justify the R&D expenditure. The federation system last year raised for the support of local community centers and for Jewish relief $1.4 billion. It's a sophisticated operation and that's why it's becoming more like the commercial sector in terms of its data gathering."
