Purpose and Methodology
For more than 30 years, these countywide, random-digit-dialed, computer-assisted telephone surveys, conducted annually during February and March, have measured systematically the continuities and changes in demographic patterns, life experiences, attitudes, and beliefs among successive representative samples of Harris County residents. Using identical items across the years, with new questions added periodically, the annual Kinder Institute Houston Area Survey has tracked America’s fourth largest city in the process of fundamental transformation.
During most of the twentieth century, Houston was essentially an Anglo-dominated, biracial Southern city, riding its location near the East Texas oil fields to continual prosperity. In May 1982, two months after the first Houston Area Survey was completed, the oil boom suddenly collapsed.
Houston recovered from deep recession in the 1980s to find itself squarely in the midst of an increasingly high-technology, knowledge-based, fully global economy and a truly remarkable transformation in its ethnic and cultural composition. New economic, educational, and environmental challenges have redefined the "pro-growth" strategies required for urban prosperity in the twenty-first century. At the same time, major immigration streams have transformed this traditionally biracial, Anglo-male-dominated, southern city into the single most ethnically diverse large metropolitan area in the country. Houston is at the forefront of the new diversity that is profoundly refashioning the social and political landscape across all of urban America.
The overall purpose of this research program is to measure systematically the way area residents are responding to these remarkable changes, to explore the bases for individual differences in attitudes and beliefs, and to make the findings of this continuing research readily available to civic and business leaders, to the general public, and to scholars everywhere.
As indicated on this site under “Survey Topics and Data,” the interviews record a rich array of socioeconomic and demographic characteristics. They measure the respondents’ perspectives on the local and national economy, on poverty programs and inter-ethnic relationships; their beliefs about discrimination and affirmative action; their attitudes toward immigration, education, crime, health care, taxation, and community service; their assessments of downtown development, mobility and transit, land-use controls, and environmental concerns; their views on abortion rights, homosexuality, and other aspects of the "social agenda." They record religious and political orientations, and a rich array of demographic characteristics, residence patterns, socioeconomic indicators, and family structures.
In order to ensure that every Harris County adult living in a household with a telephone (either land line or cell) will have an equal probability of being interviewed, the survey respondents are selected annually through a two-stage procedure. In each household reached by random digit dialing, the eligible respondent is selected randomly from all household members aged 18 or older, with initial preference given to an adult male. Using "back translation" and the reconciliation of discrepancies, each year's questionnaire is translated into Spanish, and bilingual interviewers are assigned to the project at all times.
In the early years, the sample sizes ranged from 412 to 550; since 1990, they have been set at around 650, and in 2010 and 2011 at 750. Beginning with the 2012 survey, the samples have been broadened to include more than 1,300 respondents from all nine of the counties that define the greater Houston metropolitan region (i.e., not only Harris County, but also Fort Bend, Montgomery, Galveston, Brazoria, Liberty, Waller, Chambers, and Austin counties). The distributions of the responses from Harris County residents in the 32 successive surveys on all items are presented on this site under “Survey Topics and Data.”
Response rates – the number of completed interviews in relation to all potentially eligible phone numbers – averaged 75 percent in the 1980s and have fallen to around 35 percent in the past few years. Cooperation rates – the ratio of completions to interviews plus refusals – remained for many years at approximately 80 percent, and they too have declined to about 50 percent more recently. These are disconcerting trends, but they are relatively high figures for survey research today, and the sample distributions justify continued confidence in the reliability of the data.
In addition, the survey data have been “weighted,” based on the latest census information, in order to correct for nonresponse bias and to align the data with known population parameters, such as gender, age, education, race and ethnicity, county population and density, and phone status (i.e., cell phone only, dual-frame, or land-line only). Using the variable “weight” will help to ensure that the survey findings provide the most accurate possible reflections of the actual attitudes and experiences to be found within the Harris County adult population as a whole.
From 1994 through 2012 (the one exception was 1996), the surveys were expanded with "oversample" interviews in Houston's ethnic communities. Using identical random-selection procedures and terminating after the first few questions if the respondent was not of the ethnic background required, additional interviews were conducted in each of these years to enlarge and equalize the samples of Anglo, African-American and Hispanic respondents at about 500 each. In 1995, 2002 and 2011, the research was further expanded to include large representative samples (N=500) from Houston's Asian communities as well, with one-fourth of the interviews being conducted in Vietnamese, Cantonese, Mandarin, or Korean.