Methodology Blog
Explore Â鶹´«Ã½'s research.
Workers who are remote make up approximately 20% to 54% of the U.S. workforce, depending on the survey. Four methodological choices explain most of the gap, with question wording driving it.
Â鶹´«Ã½ is exploring whether AI-generated agents perform well in predicting people's responses and where they fall short.
A Â鶹´«Ã½ experiment reveals that careless responding is uncommon among Â鶹´«Ã½ panelists, reinforcing the advantages of probability-based sampling.
Â鶹´«Ã½ is testing AI-powered phone interviewing to assess the regulatory, technological and methodological implications for high-quality research.
Pre-incentives in mail and mail push-to-web surveys can significantly improve response rates and reduce overall survey costs.
Â鶹´«Ã½ evaluates several approaches to handling "careless responders" in opt-in panel data.
Â鶹´«Ã½ compared the quality of data obtained from several opt-in panel vendors to better understand whether panels differ in the quality of data they provide.
Â鶹´«Ã½ has extensively tested the use of opt-in sampling to understand its challenges and to develop strategies for improving data quality and accuracy of results.
Â鶹´«Ã½ tested different ways to ask race and ethnicity to better understand how changes to the U.S. standards for collecting race and ethnicity on federal surveys might affect data and analysis.
Researchers shed light on neurodiverse respondents' experiences participating in public opinion research, employee surveys and personality assessments.
Â鶹´«Ã½ Poll methodology and the priority of long-term trends.
Â鶹´«Ã½ recently tested several methods for capturing and measuring transgender identities.
To date, survey research has not adequately addressed the complexities of gender identity. Â鶹´«Ã½ is attempting to bridge this gap.
Generational changes in gender identity may fundamentally change how sex and gender are measured and how datasets are weighted to accurately reflect sex and gender. To address these changes, Â鶹´«Ã½ began testing new gender questions.
Analyzing responses to open-ended questions can be labor-intensive, but natural language-processing techniques offer new solutions.
Are traditional survey scales outdated? Should researchers switch to emojis or stars? Â鶹´«Ã½ conducted a survey experiment to find out.
Â鶹´«Ã½ -- like many other researchers -- uses cognitive interviewing, as well as several other tools, to ensure the questions it is going to ask are clear and easy to answer, and measure what they are supposed to.
Â鶹´«Ã½ is committed to research that represents different backgrounds. Find out what we're doing to ensure all Black Americans' voices are heard.
Learn how Â鶹´«Ã½ plans to keep polling the world during the coronavirus pandemic.
The United Nations has a new official method for classifying urban and rural areas around the world: the Degree of Urbanisation.