Grants and Contributions:

Back to search

Title:

Bayesian Methods for Imprecise Data

Agreement Number:

RGPIN

Agreement Value:

$70,000.00

Agreement Date:

May 10, 2017 -

Organization:

Natural Sciences and Engineering Research Council of Canada

Location:

Newfoundland and Labrador, CA

Reference Number:

GC-2017-Q1-01862

Agreement Type:

Grant

Report Type:

Grants and Contributions

Additional Information:

Grant or Award spanning more than one fiscal year. (2017-2018 to 2022-2023)

Recipient's Legal Name:

Abarin, Taraneh (Memorial University of Newfoundland)

Program:

Discovery Grants Program - Individual

Program Purpose:

In statistical association studies, outcomes are determined by combinations of error-prone and accurately measured variables that interact with each other. Association models with interaction terms are widely applied in many areas, such as genetics, engineering, economics, education and epidemiology. For example, an interaction model is considered to assess whether or not the magnitude of an error-prone variable (such as self-reported average number of cigarettes smoked a day) on an outcome (e.g. risk of heart attack) is modified by accurately measured variables (e.g. age, gender, presence or absence of family history of a heart attack). As individuals in the study may have imprecise recall, the variable “average number of cigarettes smoked a day” is considered imprecise or measured with error. In order to improve the accuracy in the assessment of the effects of these variables on the outcome, one needs to take into account these errors. As techniques dealing with interaction terms when error-prone variables are involved, are quite challenging, in practice, we often see the use of additive models that ignore the interaction effects. However, erroneously omitting interactions in those models significantly reduces efficiency of these studies.
There are only a few studies that recently implemented Bayesian (an advanced statistical method based on pre-experimental knowledge about unknown parameters) techniques into interaction models with measurement error. However, these studies are still at the early stage of development with many gaps to be filled. For example, majority of them consider models that do not include multiple accurately measured variables, which is more realistic in application. The proposed research is going to fill some of these gaps. It focuses on general association studies with the primary focus on interaction terms subject to error, within the Bayesian paradigm. The findings of this program will shed some light on some of the complex challenges in association studies.