Large surveys are becoming more expensive and time-consuming with substantial survey burden, creating significant non-responses, thereby making the surveys unreliable. Administrative data, organic data and data from web-scraping can overcome these issues, not by themselves but with combination with smaller surveys. Currently, this is an area of enormous interest as survey organizations thrive to reduce cost and improve precision. In this invited session, renowned researchers (from different geographic regions and ethnicity) will share the latest advances developed in this field. Therefore, this invited session can attract a relatively large number of attendees at the 2023 Ottawa Conference.
 

There are three presenters and one discussant.  The presenters are Dr. Balgobin Nandram (Worcester Polytechnic Institute, USA), Dr. Serena Arima (Università del Salento, Italy), Dr. Yajuan Si (University of Michigan, USA), and the discussant is Dr. J.N.K. Rao (Carleton University, Canada). The speakers are renowned researchers in the filed of survey sampling. The speakers have carefully been selected to also represent different geographic regions and also diversity based on the ISI mission (equity, diversity, inclusion). Each speaker will talk on a different aspect of data (e.g., administrative, survey) integration, and Dr. Rao will summarize the talks with giving some suggestions/comments and directions for further research in this field. 

In particular, Dr. Nandram will talk on Bayesian data integration for predictive inference about small areas where a relatively small probability sample and a non-probability sample are available from each area. He will show that the data-integrated model provides small area estimates, mostly similar to those of the probability sample only model, but with larger precision. Dr. Arima will talk on modeling of misreported data which come from combination of registry and survey data. The aim is to develop statistical tools for analyzing misreported data, a problem that frequently occurs when dealing with social science data, note that misreporting is typically treated as a nuisance factor to be removed from the analysis with ex-post correction methods, rather than to be considered as a structural component in the model specification. Dr. Si will talk on sampling design for multiple surveys. The aim is to develop strategies for a coordinated sampling process to potentially reduce burdens on decision-makers and respondents, decrease refusal/nonresponse, and exploit the variables collected across surveys to leverage data from multiple surveys.

Organiser: Prof. Mahmoud Torabi 

Chair: Prof. Mahmoud Torabi 

Speaker: Dr Yajuan Si  

Speaker: Prof. Balgobin Nandram 

Speaker: Prof. Serena Arima 

Discussant:  J.N.K. Rao