Survey methodology

The 2000 Iowa Child and Family Household Health Survey was a telephone interview conducted with a stratified random sample of 3,200 families with children in Iowa. The interview included approximately 125 questions, depending on the number of questions relevant to the family being interviewed. The research team developed the survey instrument after reviewing existing instruments such as the National Survey of American Families (NSAF) and the National Health Interview Survey (NHIS).1,2

To identify families with children in Iowa, researchers began with a random list of phone numbers provided by a private vendor. To allow for regional comparisons, 400 interviews were completed in each of eight regions of the state. All regions encompassed multiple counties with two exceptions—Polk and Scott counties were single-county regions. The design of this study yielded a representative sample of families with children in Iowa. To make statewide estimates, the survey results were weighted and post-stratified using the county-level 2000 US Census data.3
The survey process began with a screening question to determine if the number dialed reached the home of a family with children. If it had, the adult most knowledgeable about the health and health care of a randomly selected child under age 18 in the household was asked to complete the interview.

The dispositions of calls made to complete the 3200 interviews were as follows:

Table 1. Participation Rate for 2000 Iowa Child and Family Household Health Survey
Number completing the interview	3241
Number of refusals or unable to complete interview	1349
Participation Rate	71%

 

Identification of race/ethnicity and extra sampling of minority children

In this study, each child's race/ethnicity was classified based on the parent/guardian's response to several questions that were similar to those used in the 2000 US Census. These questions were designed to partially distinguish between race and ethnicity, especially for Latinos.

Question 1: Is your child of Spanish or Hispanic Origin?
Question 2: What is your child's race?

[Respondents could report more than one race. Responses were coded—African-American, White, American Indian/Native American/Aleutian or Eskimo, Asian/Pacific Islander, or Other].

To ensure that the sample would include enough minority children to allow for comparisons of the results by race/ethnicity, an additional 457 households with minority children were called as part of an oversample after the initial 3,200 interviews were completed. These calls were targeted to areas of the state with higher proportions of minority families. The total number of respondents by race was:

• African-American: 134
• Asian: 98
• Latino: 379
• White: 2,932
• More than one race: 84
• Other: 99

The telephone interviews were conducted between May and October 2000 by the Center for Social and Behavioral Research at the University of Northern Iowa. The University of Northern Iowa Human Subjects review board approved the protocol regarding the telephone interview portion of this study. The survey was translated into Spanish by the Translation Laboratory at the University of Iowa. One Spanish-speaking interviewer was employed by the University of Northern Iowa Center for Social and Behavioral Research to conduct all of the interviews in which parents preferred speaking Spanish. Forty-three interviews were completed in Spanish and are included in these analyses. In any telephone-based survey, there is a possibility that results may be biased because those without telephones are not interviewed; people without telephones may have different health conditions and health care needs than those with telephones. In Iowa, it is estimated that three percent of households do not have telephones.

Comparison of results by race/ethnicity
Comparisons were made for children in racial groups that had about 90 or more responses (before weighting). For this report, results are presented for children who were reported by their parents as African-American, Latino, Asian or White. These groups were compared on several factors including:

• demographics
• functional health status
• health insurance coverage
• access to care and use of services
• child and family well-being, including school performance, child care and the family environment