Abstract
High-frequency social data collection may facilitate improved recall, more inclusive reporting, and improved capture of intra-period variability. Although there are examples of small studies collecting particular variables at high frequency in the social science literature, to date there have been no significant efforts to collect a wide range of variables with high frequency. We have implemented the first such effort with a smartphone-based data collection approach, systematically varying the frequency of survey task and recall period, allowing the analysis of the relative merit of high-frequency data collection for different key variables in household surveys. This study of 480 farmers from northwestern Bangladesh over approximately one year of continuous data on key measures of household and community wellbeing could be particularly useful for the design and evaluation of development interventions and policies. While the data discussed here provide a snapshot of what is possible, we also highlight their strength for providing opportunities for interdisciplinary research in the household agricultural production, practices, seasonal hunger, etc., in a low-income agrarian society.
Highlights
Background & SummaryConventional household surveys typically use multiple visits with a large time gap between visits to construct longitudinal data[1]
We expect the dataset to provide a high-frequency, intra-annually resolved window into the dynamics of topics commonly summarized as single numbers in an integrated household survey, albeit for a small and purposively selected sample
Through our local implementing partner WIN Incorporated we assigned four local Points of Contact (POC) in each upazilla, who consulted with local agricultural extension officers to make a list of technology early adopting farmers, who the officers deemed likely to be among the first to engage with smartphones
Summary
Conventional household surveys typically use multiple visits with a large time gap between visits to construct longitudinal data[1] These data typically suffer from recall bias and lost intra-period variation, as respondents are asked in one sitting to recall events or outcomes that have transpired during the entire period between survey interviews[2,3]. By putting the smartphones directly into the respondents’ hands, our study could capture data from those at the focus of study, even those in remote places where access is limited or unreliable This dataset contains approximately one year’s worth of high frequency data containing information on a wide variety of experiences, inputs, and outcomes such as agricultural production and practices; experiences with. We expect the dataset to provide a high-frequency, intra-annually resolved window into the dynamics of topics commonly summarized as single numbers in an integrated household survey, albeit for a small and purposively selected sample
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
