Demographics of Public and Doctors Surveyed
Markle Health in a Networked Life
The Markle Survey on Health in a Networked Life uniquely compares the core values of physicians and patients on deployment of information technology in health care.
- Nearly one-quarter of the public say they’ve had trouble paying medical bills.
- Very few report privacy as a reason to pay own expenses or avoid care.
General Population Sample
An initial sample of 2,493 people was drawn randomly from KN's KnowledgePanel, a probability-based web panel designed to be representative of the US non-institutionalized population age 18 and up, including both online and offline households (households are provided access to the Internet and hardware if needed). From the initial sample, 1,582 responded to this online survey—a completion rate of 63 percent. The survey was conducted August 10 through August 24, 2010.
An initial sample of 2,867 physicians was drawn from Knowledge Networks Physicians Consulting Network, an online opt-in convenience panel consisting of physician members invited to join from reputable listed samples, such as the AMA Masterfile and prescriber lists. From the initial sample, 779 physicians responded to this online survey—a completion rate of 27 percent. The physician survey was conducted August 10 through August 26, 2010. The panel generally reflects the US physician population on known demographics, although it somewhat under-represents younger physicians and hospital-based physicians. The study sample consisted of 26 percent primary care doctors (defined as internal medicine, general practice, or family practice) and 74 percent specialists.
The general population sample can validly be projected to the US population, with margin of error of ±3.0 percent. Although we believe the physician sample is a highly defensible reflection of US physicians, because it was not drawn entirely by random we are not as comfortable projecting its results to all US physicians. Because the sample is based on those who were invited to join from multiple sources without known probabilities of selection, estimates of theoretical sampling error cannot be calculated.