March 16, 2025
Education News Canada

UNIVERSITY OF WINNIPEG
Improving accuracy in statistical data

March 13, 2025

The use of accurate statistics is integral in any informed decision or policy making, in business, government, university, or household. University of Winnipeg's is a statistician who has a passion for improving the value of data collected using the science of statistics to help improve its accuracy.


Dr. Zeinab Mashreghi
 

Dr. Mashreghi grew up in a family passionate about mathematical sciences, which shaped her interest in both pure and applied mathematics. Her academic journey began with applied mathematics in her Bachelor of Science; she then moved to pure mathematics in her Master's of Science; concluding in statistics in her PhD. This is where she found the mathematical sweet spot - the perfect balance between theory and application.

"The ability to apply mathematical principles to data and solve real-world problems is what makes statistics so fascinating to me," shared Dr. Mashreghi.

Dr. Mashreghi's research enhances statistical analysis helping national statistical offices like Statistics Canada, and researchers in various fields, including  social and health sciences with access to detailed datasets to produce more precise survey results.

Ultimately, my goal is to contribute to more precise and resilient research results for crucial informed decision-making in any field of study.

Zeinab Mashreghi

To further her research, Dr. Mashreghi was given a Discovery Development Grant for her research project, Leveraging Bootstrap Methods for Enhanced Estimation in High-dimensional Data and Addressing Missing Survey Data, valued at $44,000.

She is working on improving the reliability of results extracted from datasets that have already been collected. Her research focuses on developing and improving statistical methods to obtain more accurate results from survey data.

Statistical surveys often face the issue of nonresponse, where some questions remain unanswered or a sampled unit refuses to participate. Dr. Mashreghi explains that if missing values are not properly addressed, estimates can become biased, leading to inaccurate conclusions. This issue becomes even more complex in high-dimensional survey datasets.

To address these challenges, she develops bootstrap resampling techniques to improve variance estimation a key measure of accuracy for a statistic ensuring more reliable results even when dealing with missing data, and the complexities of high-dimensional survey structures.

"The methods I develop are useful for statisticians and researchers across various fields who analyze survey data," said Dr. Mashreghi. "National statistical agencies like Statistics Canada, can apply these techniques to improve the accuracy of their survey results."

Dr. Mashreghi is also developing R packages to enable researchers across various fields to easily apply bootstrap techniques to survey data, helping them improve the reliability of their research conclusions.

These packages will make both existing and newly developed bootstrap methods more accessible to researchers and graduate students, helping them apply advanced statistical techniques in their studies and further advance research in their fields.

"Ultimately, my goal is to contribute to more precise and resilient research results for crucial informed decision-making in any field of study," said Dr. Mashreghi.

For more information

University of Winnipeg
515 Portage Avenue
Winnipeg Manitoba
Canada R3B 2E9
www.uwinnipeg.ca/


From the same organization :
14 Press releases