Hispanic America Study

Hispanics represent nearly one out of every six adults in the U.S. today and demographic projections show this segment is likely to grow in the coming decades.

By 2050, it’s predicted there will be no ethnic or racial majority in the U.S. and Hispanics will make up 25 to 30 percent of the population.

In the coming decades, the values, beliefs, behaviors and worldview of Hispanics will increasingly affect the fabric of American life. Understanding the Hispanic audience and their impact is important for anyone interested in reading the times and carefully engaging with cultural and sociological trends.

It is with an eye toward these shifting demographics that Barna Group, in partnership with the American Bible Society, the National Hispanic Christian Leadership Conference and OneHope, conducted the study Hispanic America: The Faith Values and Priorities of a Rapidly Growing Audience. The study explores various aspects of Hispanic’s faith, social attitudes and lifestyles. The study included 2,046 interviews nationwide with individuals who are Hispanic or Latino. The central objective of the research is to provide a snapshot of today’s Hispanic audience, with a view toward the influence of the faith and the Bible on real-life issues facing Hispanics.

You can purchase a full version of the study here.

Survey Methodology

The data contained in the Hispanic America study originated through a research project conducted by the Barna Group of Ventura, California. The study was commissioned by Barna: Hispanics, the American Bible Society, OneHope, and National Hispanic Christian Leadership Conference.

A total of 2,046 interviews were conducted among Hispanic adults living in the United States; 1,044 online and 1,002 by phone in August and September 2012. The sampling error for 2,046 interviews is +/-2.2 percentage points, at the 95% confidence level. All interviews were offered in English and Spanish. All interviews were with one household member only selected at random.

Oversampling of Protestants was conducted to help represent this important part of the emerging profile of Hispanic faith, and to facilitate comparisons among groups with different religious affiliations. The final sample was then weighted back to the natural, nationwide distribution of Catholics and Protestants, and other religious affiliations. At the start of interviewing, all self-identified Hispanics were engaged.  In order to create an oversample of Protestants, during the course of interviewing, only self-identified Hispanics who also identified themselves as Protestants were allowed to continue.

Hispanic Segmentation

  • Catholic: sample size, n=1,072
  • Protestant: sample size, n=592
  • Other faith / no faith: sample size, n=382

Telephone Interviews

The telephone survey included 1,002 interviews conducted among a representative sample of Hispanic adults, 18 years of age and older, from within the 48 continental states. The survey was conducted from August 16, 2012 through September 4, 2012. The average interview length was 16 minutes. The sampling error for the phone sample on its own is +/-3.1 percentage points, at the 95% confidence level. All interviews were conducted with live interviewers.

Most interviews were conducted via respondents’ home telephones (a.k.a., landlines). In this study, 40% of the total interviews in the study were conducted with respondents on their cell, mobile, or smartphones. This was done to ensure that households that only use a cell or mobile were included.

Telephone interviews were conducted by bilingual interviewers—respondents were asked if they would prefer to take the survey in English or Spanish. Most respondents (76%) chose to be interviewed in English, and the remainder (24%) chose Spanish. All telephone interviews for the project were conducted with experienced, trained interviewers. The interviewers were supervised at all times, as well as monitored during the course of their work on this project. The survey was conducted through the use of a CATI (Computer Assisted Telephone Interviewing) system. This process ensures that question skip patterns are properly administered by interviewers and that survey data are recorded accurately.

The survey calls were made at various times during the evening and weekend so that every individual selected for inclusion in the research was contacted numerous times on separate days, at different times of day, to maximize the possibility of contact. This is a quality control procedure that ensures individuals on the sample list have an equivalent probability of inclusion within the survey, thereby increasing the survey reliability.

The sample of 602 landline telephone exchanges called was randomly selected by a computer from a complete list of tens of thousands of active residential exchanges across the country. The area codes and their exchanges were chosen at random, but proportionally to the population to ensure that each region of the country was represented in proportion to its share of all telephone numbers. These samples were dialed using Computer Assisted Telephone Interviewing Stations. Within each exchange, random digits were added to form a complete telephone number, thus permitting access to listed and unlisted numbers alike.

To increase coverage, this landline sample was supplemented by respondents reached through random dialing of cell phone numbers. A total of 400 completes were obtained through cell phone samples. The two samples were then combined and adjusted to assure the proper ratio of landline-only, cell phone-only, and dual phone users. Cell phone samples were dialed by hand in accordance with current law.

Interviewers made no fewer than 8 attempts over the course of fielding in order to reach someone at every phone number in the survey, calling back unanswered numbers on different days at different times of the day and evening until someone answered, participated or were deemed not qualified for this survey.

Online Interviews

The Hispanic study also included 1,044 online surveys conducted among a representative sample of Hispanic adults, ages 18 and older, throughout the United States. Interviews were conducted from August 10, 2012 through August 22, 2012. The average survey length was 19 minutes. The sampling error for the online sample alone is +/-3.1 percentage points, at the 95% confidence level.

Respondents were given the option of taking the survey in either English or in Spanish. Most (63%) chose to take the survey in English, and the remainder (37%) took it in Spanish.

This study used an online research panel from GfK Knowledge Networks. This high-quality panel is based on probability sampling that covers both the online and offline populations in the U.S. The panel members are randomly recruited either by telephone (Random Digit Dialing) or by self-administered mail and web surveys (Address-Based Sampling). Households are provided with access to the Internet and hardware if needed. Unlike other Internet research that covers only individuals with Internet access who volunteer for research, this process uses a dual sampling frame that includes both listed and unlisted phone numbers, telephone and non-telephone households, and cell-phone-only households. The panel is not limited to current Web users or computer owners. All potential panelists are randomly selected to join the KnowledgePanel; unselected volunteers are not able to join.


The combined results have been weighted to adjust for variation in the sample relating to geographic region, sex, race, marital status, age, education and the number of adults in the household.  Protestants were also weighted to their natural distribution among Hispanics. Those statistical weights were designed and applied from the United States Census Bureau statistics.

Respondents were weighted both within religious categories and across religious categories using the following four-stage procedure:

For Stage One, respondents were weighted to look like the 18+ Hispanic U.S. population on the standard weighting variables.  Additionally, they were also weighted on primary language—since the sample includes both English and Spanish survey takers—and Hispanic origin (Mexican, Puerto Rican, Cuban, other).  The weights were then trimmed and scaled to sum to the sample size of total qualified respondents.

For Stage Two, the un-trimmed post-stratification weights from the first-stage weights were used to set weighting benchmarks within each religion group: Catholic, Christian/Non-Catholic, and other.

For Stage Three, respondents were weighted to look like the benchmarks from Stage Two by controlling the demographics within the three religion groups (Catholic, Christian/Non-Catholic, other) on the standard weighting variables, primary language, and Hispanic origin. Weights were again trimmed and scaled to sum to the sample size of total qualified respondents.

Finally, weights were collapsed across the weighting variables as follows:

  • Catholic: Gender, Age, Region, Metro, Education, Internet, Household Income, Primary Language, Hispanic Origin (Mexican, Puerto Rican, Cuban, all others)
  • Christian/Non-Catholic: Gender, Age, Education, Internet, Household Income, Primary Language, Hispanic Origin (Mexican, Puerto Rican/Cuban, all others)
  • Other Faith: Gender, Age, Education, Internet, Household Income, Primary Language, Hispanic Origin (Mexican, non-Mexican)
  • Overall: Gender by Age, Region, Metro, Hispanic Origin (Mexican, Puerto Rican, Cuban, all others)