Survey Methods

You are here

The goal of the NPCH is to assess issues in a timely fashion using nationally representative scientific probability samples of U.S. households. Periodic surveys are conducted using an innovative, rigorous, established Web-based survey technology provided by a private vendor (GfK Custom Research, LLC). The advantage of the novel Web-based survey mechanism is that surveys dedicated to specific issues (such as adolescent health) can be targeted to households for whom the questions are most relevant (such as households with adolescents). In addition, the field period for Web-based surveys is brief (2-4 weeks). 

Our surveys address a wide variety of children's health issues, including child health conditions (such as asthma and obesity), insurance coverage, preventive services, access to care, children's special health needs, relationship of health and education for children, and safety and health concerns. In all cases, adults are the respondents and their responses are anonymous. For each survey, there are approximately 2000 respondents – in households with children and households without. GfK Custom Research, LLC (GfK) has established the first online research panel based on probability sampling that covers both the online and offline populations in the U.S., called KnowledgePanel®. The panel members are randomly recruited by random digit dialing and address-based sampling techniques. Households are provided with access to the Internet and a laptop if needed. Unlike other Internet research that covers only individuals with Internet access who volunteer for research, KnowledgePanel® surveys are based on a sampling frame that includes both listed and unlisted phone numbers, and is not limited to current Web users or computer owners. Panelists are selected by chance to join the panel; unselected volunteers are not able to join the KnowledgePanel®.

All National Poll on Children's Health survey samples are drawn from the KnowledgePanel®. To reduce the effects of potential nonresponse and noncoverage, GfK calculates panel weights using national demographic distributions, as per the most recent Current Population Survey. These include age, race, gender, Hispanic ethnicity, and education, and are applied prior to the selection of a sample. In addition, the National Poll on Children's Health sample design oversamples households with parents so that more precise estimates can be made of parent opinions. Survey sample data are submitted to a post-stratification adjustment once they return from the field. Iterative proportional fitting is used to reduce sampling variance for characteristics highly correlated with population benchmarks, and to reduce bias due to nonresponse. By using these methods, generalizations can be made from these survey samples to the national U.S. population of parents and to the U.S. population of all adults.

Analysis of the survey data includes examining relationships among survey items, between survey items and external datasets and programs, such as the Center for Disease Control and Prevention's growth charts, or between survey items and respondents’ demographic data provided by the vendor. Demographics such as gender, age, race, income, and education are collected in a follow-up survey for each panel member to create a member profile. This information can be used to determine eligibility for specific studies and need not be gathered with each survey. Once this survey is completed, the panel member is regarded as active and ready to be sampled for other surveys.