Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Discovery Grants Program - Individual

Simon Fraser University

$28,000.00

Knowledge Discovery Using Nonrandom Data Samples

EF 2024-25

Résumé
Real world non-random data such as observational data are widely available and offer opportunities for better decision making and evidence-based innovations. Advancements in digital technology have generated an increasing amount of rich observational data sets, such as large administrative databases, and intensive electronic diary/survey data collected through real-time mobile devices. Even randomized experiments can yield non-random data, such as real world randomized control trials (RCTs) that are frequently plagued by intervention noncompliance. Although these non-random data offer tremendous opportunities to gain real word evidence on risk factors, social determinants, and intervention effectiveness, significant bias can be introduced into standard statistical inference due to various common data limitations, notably non-random sampling and retention of study participants, non-random receipt of treatment and exposure variables. Analytical methods relying on the tacit but crucial assumption of randomness can yield severely biased results that have poor external generalizability and low internal validity. In response to the strong demand for appropriate analytic techniques for use with these non-random data, my research program aims to develop novel and principled statistical methods  that are flexible and feasible to use for proper knowledge discovery in today's rich data environment. The proposed research will build on the recent progress of the applicant, which includes new bias controlling methods for non-random data that have been published in statistics and quantitative science journals, and have had an impact in domain-specific applied areas. The short-term objectives are: A) novel causal inference methods for experimental data with multiple endpoints and intervention noncompliance, B) novel instrument-free joint estimation methods for causal inference using observational data, and c) novel selection bias control methods to handle non-randomly sampled/missing data. Longer-term goals include: 1) develop general, flexible, robust, and tractable statistical methods that increase the usability, reliability, and validity of non-random data for knowledge discovery, 2) provide excellent training to new statisticians, and 3) educate practitioners on proper use of novel statistical methods and facilitate the uptake of these methods via open-source software programs. The methodological approach will include analytical derivations for both simple and generalized models, a study of performance through both computer simulation experiments and applications to real data sets. The research program will provide ample opportunities for trainees to get involved in all aspects of the research and gain knowledge in developing and applying statistical methods to diverse, new data sets emerging from across sectors. The proposed work will advance new analytical methods and improve the efficiency, precision or speed of data analysis by applied researchers.
Programme
Discovery Grants Program - Individual
Emplacement
British Columbia, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data

Tel que publié

Application id
RGPIN-2024-06629
Name nom
XIE, HUI
Department département
Faculty of Health Sciences
Organization id
5
Institution établissement
Simon Fraser University
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
28000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1508
Committee name en
Mathematics and Statistics
Committee name fr
Mathématiques et statistique
Area of application code
1100
Area of application group en
Health, education and social services
Area of application group fr
Santé publique, éducation et services sociaux
Area of application en
Health, education and social services
Area of application fr
Santé publique, éducation et services sociaux
Research subject code
3005
Research subject group en
Statistics and probability
Research subject group fr
Statistique et probabilité
Research subject en
Biostatistics
Research subject fr
Biostatistiques
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Knowledge Discovery Using Nonrandom Data Samples
Keywords
selection bias; endogeneity; unobserved confounding; omitted variable bias; missing data; causal modeling; longitudinal data; sensitivity analysis; confounding bias; nonignorable
Application summary
Real world non-random data such as observational data are widely available and offer opportunities for better decision making and evidence-based innovations. Advancements in digital technology have generated an increasing amount of rich observational data sets, such as large administrative databases, and intensive electronic diary/survey data collected through real-time mobile devices. Even randomized experiments can yield non-random data, such as real world randomized control trials (RCTs) that are frequently plagued by intervention noncompliance. Although these non-random data offer tremendous opportunities to gain real word evidence on risk factors, social determinants, and intervention effectiveness, significant bias can be introduced into standard statistical inference due to various common data limitations, notably non-random sampling and retention of study participants, non-random receipt of treatment and exposure variables. Analytical methods relying on the tacit but crucial assumption of randomness can yield severely biased results that have poor external generalizability and low internal validity. In response to the strong demand for appropriate analytic techniques for use with these non-random data, my research program aims to develop novel and principled statistical methods  that are flexible and feasible to use for proper knowledge discovery in today's rich data environment. The proposed research will build on the recent progress of the applicant, which includes new bias controlling methods for non-random data that have been published in statistics and quantitative science journals, and have had an impact in domain-specific applied areas. The short-term objectives are: A) novel causal inference methods for experimental data with multiple endpoints and intervention noncompliance, B) novel instrument-free joint estimation methods for causal inference using observational data, and c) novel selection bias control methods to handle non-randomly sampled/missing data. Longer-term goals include: 1) develop general, flexible, robust, and tractable statistical methods that increase the usability, reliability, and validity of non-random data for knowledge discovery, 2) provide excellent training to new statisticians, and 3) educate practitioners on proper use of novel statistical methods and facilitate the uptake of these methods via open-source software programs. The methodological approach will include analytical derivations for both simple and generalized models, a study of performance through both computer simulation experiments and applications to real data sets. The research program will provide ample opportunities for trainees to get involved in all aspects of the research and gain knowledge in developing and applying statistical methods to diverse, new data sets emerging from across sectors. The proposed work will advance new analytical methods and improve the efficiency, precision or speed of data analysis by applied researchers.