Research and Development Survey (RANDS)

The Research and Development Survey (RANDS) examines methods used to detect measurement error and to calibrate survey method estimates.  The specific questions vary by round, but the survey is always focused on health behaviors and conditions.  Topics that have been included on RANDS include

• Self-rated health
• Chronic conditions including diabetes, hypertension, lung conditions, and high cholesterol
• Smoking behaviors
• Drinking behaviors
• Physical activity
• Affect and psychological distress
• Pain
• Opioid use

The RANDS questionnaires are primarily composed of National Health Interview Survey questions, as well as some NCHS-developed probes that are designed to capture the response process.  These questionnaires have been administered to members of commercially available, recruited, “probability-based” web panels.  Each round of RANDS has approximately 2,000 respondents.

Additionally, NCHS adapted RANDS to collect timely information on the COVID-19 pandemic.  This special iteration of RANDS, named RANDS during COVID-19 to distinguish it from previous RANDS surveys, was conducted across two rounds using both a probability-base and opt-in commercial survey panel and featured larger sample sizes than previous rounds of RANDS (approximately 6,000 respondents from the probability panel and 10,000 respondents from the opt-in panel, per round).

RANDS during COVID-19 was conducted to: a) evaluate new COVID-19 related survey questions being fielded on traditional surveys, such as the National Health Interview Survey; b) to conduct survey methods analysis of probability and opt-in sample platforms and internet and telephone survey modes; and, c) to compare selected COVID-19 related constructs with those from other federal data collections, including the Household Pulse Survey. RANDS methodological surveys, including RANDS during COVID-19, have not been evaluated for other statistical purposes, including health outcomes research and model-based association studies. In addition to addressing methodological issues, a selected set of questions unavailable from other sources were added to RANDS during COVID-19 to take advantage of the platform to provide experimental estimates on topics of policy interest including telemedicine, reduced access to care due to the pandemic, and work loss.  Limitations of the experimental estimates are described in the technical notes:

Data and Documentation:
Round 1- Autumn 2015 pdf icon[PDF – 58 KB]
Round 2- Spring 2016 pdf icon[PDF – 79 KB]
Round 3- Spring 2019 pdf icon[PDF – 191 KB]
Round 4- Summer 2020 pdf icon[PDF – 233 KB]
RANDS during COVID-19 Round 1 Non-Probability Sample pdf icon[PDF – 110 KB]
RANDS during COVID-19 Round 1 Probability Sample pdf icon[PDF – 116 KB]
RANDS during COVID-19 Round 2 Non-Probability Sample pdf icon[PDF – 117 KB]
RANDS during COVID-19 Round 2 Probability Sample pdf icon[PDF – 124 KB]

Each file includes questionnaire data, limited paradata, some demographic and characteristic data, and sample weights.  The datafiles include records on both survey responders, partial responders, and sampled non-responders



Page last reviewed: January 7, 2021, 02:35 PM