ASA/NSF/BLS Fellowship Possible Research Topics in Behavioral Science
main Fellowship page for more information on eligibility and application requirements. Applicants may also be interested in topics in Statistics or Economics.
Behavioral science research approaches help to better understand psychological, sociological, and anthropological
factors in the survey data collection process. Research in this area could be
conducted with existing data or with studies conducted in the field or in the BLS Behavioral Science
Research Laboratory. The lab is a fully equipped usability lab and observation room, with audio and video
recording capabilities, and facilities for running computer-administered
Many BLS surveys utilize interviewers for data collection. Research shows that interviewers can significantly impact respondents, the interview process, and the data collected. Obtaining a better understanding of the interactions between interviewers and respondents is important as BLS develops training, instruments, and procedures. Potential research topics include:
- Coding recordings of interviews to understand how interviewers follow scripted questions and procedures (or not)
- Exploring the impact of non-scripted probes on respondent responses
- Analysis of survey paradata (e.g., audit trails, CARI recordings) to evaluate interviewer performance and data quality
Nearly all BLS household surveys rely on respondentsí abilities to recall information about their past behaviors to answer survey questions. Having a good understanding of how autobiographical information is stored and retrieved from memory will help BLS design more effective survey questions and procedures. Potential research topics include:
- Identifying features of the survey (e.g., question topic or format, reference period, interviewing protocols, etc.) and respondent (e.g., cognitive ability, motivation, etc.) that impact retrieval and reporting of autobiographical information
- Developing optimal retrieval strategies and cues for reporting quantitative aspects of autobiographical events (e.g., expenditures, activity durations, frequency estimates, etc.)
Like most surveys, achieving or maintaining a high response rates for BLS establishment surveys has become more difficult over time. We seek to better understand non-response across our establishment surveys, answering questions such as:
- Are there establishment characteristics related to nonresponse?
- Are there certain response patterns in panel surveys that are related to nonresponse?
- Are there characteristics of the survey or questions that are related to future nonresponse?
As the use of technology advances, BLS programs are in the process of adding new modes to existing surveys to accommodate respondent preferences. Examples include the web diary in the American Time Use Survey, and use of mobile devices in the Consumer Expenditure Survey. There are many issues related to this, including:
- Understanding mode effects, such as the relationships between mode, response, and data quality
- Understanding how to introduce multiple modes to respondents, in parallel or sequentially
- Exploring how to combine data collected using different modes (e.g., whether post-collection adjustment is necessary)
BLS designs and builds many kinds of systems, and we are always looking to improve the user experience of our products. Potential research topics include:
- Designing paper and web forms to better meet respondentsí needs and expectations
- Improving methods for finding data on the BLS website
- Exploring graphical alternatives for displaying data
BLS strives to provide accessible products. Potential relevant research topics in this area include:
- Understanding issues related to the accessibility of our data collection instruments and public website
- Designing complex data tables that are fully accessible
Last Modified Date: November 21, 2014