Survey Methods

We field National Poll on Healthy Aging (NPHA) surveys twice per year online and by phone to a randomly selected, stratified group of U.S. adults age 50 and older drawn from NORC’s AmeriSpeak® panel, a nationally representative probability-based panel of U.S. households. We then process the data for quality and weight the data to national Census benchmarks.

We have made some changes to our data collection and sampling over time. NPHA Waves 1-8 (April 2017-January 2021) went to participants in the Ipsos KnowledgePanel®. In addition, NPHA Waves 1-13 (April 2017-August 2023) include data from surveys that went to U.S. adults age 50 to 80. Beginning with NPHA Wave 14 in 2024, we expanded the respondent age range to include those over age 80.

Starting in August 2023, we also field Michigan Poll on Healthy Aging surveys twice per year online and by phone alongside the NPHA to a sample of Michigan adults ages 50 and older drawn from NORC’s AmeriSpeak® panel as well as non-probability samples. We combine the probability and non-probability samples and use TrueNorth®, an advanced statistical technique that calibrates the combined samples to both statewide Census population benchmarks and estimates from small area models to accurately reflect Michigan population figures from the U.S. Census and reduce known biases in non-probability samples.

The webpage for each National Poll on Healthy Aging and Michigan Poll on Healthy Aging report includes fielding-specific methodology such as poll questions and response options, fielding dates, sample sizes, oversampling criteria, survey completion rate, and margin of error.

These polls use robust methods to provide valuable insights into the health-related experiences of adults age 50 and older around the nation and in Michigan. Yet, it is incumbent on us to note that there are no perfect studies, and all research and methods have their limitations. These polls are subject to limitations inherent to survey research. For example, self-reported data introduces the possibility of response biases, such as social desirability bias and recall bias, which may affect the accuracy of the information collected. Our surveys exclude people living in congregate settings such as nursing homes, which means we cannot generalize our findings to all older adults. In some cases, we are unable to provide estimates for some subgroups due to small sample sizes. We share this to make apparent some possible limitations and known sources of measurable or estimated error. While we actively take steps to minimize these limitations, there will be some unmeasured and unknowable error with all forms of public opinion research, including ours.

The University of Michigan Institute for Healthcare Policy and Innovation (IHPI) directs the National Poll on Healthy Aging, which receives support from Michigan Medicine, U-M's academic medical center. AARP also supported the poll from 2017 to 2025. IHPI also directs the Michigan Poll on Healthy Aging, which is supported by the Michigan Health Endowment Fund and Michigan Medicine, U-M's academic medical center.

Findings from the National Poll on Healthy Aging and Michigan Poll on Healthy Aging do not represent the opinions of the University of Michigan or its funders. The University of Michigan reserves all rights over this material.