4 classes matched your search criteria.

PUBH 8142 is also offered in Spring 2019

PUBH 8142 is also offered in Spring 2016

PUBH 8142 is also offered in Spring 2014

PUBH 8142 is also offered in Spring 2013

## Spring 2019 | PUBH 8142 Section 001: Epidemiologic Uncertainty Analysis (63266)

- Instructor(s)
- Class Component:
- Lecture
- Credits:
- 2 Credits
- Grading Basis:
- S-N only
- Instructor Consent:
- No Special Consent Required
- Instruction Mode:
- In Person Term Based
- Times and Locations:
- Regular Academic Session01/22/2019 - 05/06/2019Tue 12:00AM - 12:00AMUMTC, East Bank
- Also Offered:

- Course Catalog Description:
- Scientific interpretation of statistical analysis as dependent on both data and assumptions. Techniques that enable an investigator to incorporate uncertainty about assumptions into a quantitative analysis. prereq: 8140
- Class Description:
- An observed relative risk (RRobs) can be described mathematically as the product of the causal relative risk (RRcausal), a desired effect measure for etiologic epidemiologic studies, and error factors for the impact on study results of imperfections in the design, conduct and analysis of the study (uncontrolled confounding, losses-to-followup, nonrandom subject sampling, subject nonresponse, missing data, exposure and disease measurement error, unjustified statistical model assumptions, and random error). When viewed from this perspective, it becomes clear that RRcausal is not identifiable (i.e., cannot be validly estimated) without making assumptions about the values of the error-factor and random-error terms. A standard quantitative analysis does not account for most study imperfections. It therefore implicitly assumes that the product of the error factors equals 1.0. This standard-practice assumption, however, has neither theoretical nor empirical justification. We therefore advise epidemiologists to replace the standard assumption with more justifiable assumptions about the values of the error-factor terms. These more-justifiable assumptions can be incorporated into a quantitative analysis with uncertainty analysis (also known as bias modeling, probabilistic sensitivity analysis, Monte Carlo sensitivity analysis). We discuss this technique in this class.
- Textbooks:
- http://www.bookstores.umn.edu/buybooks.cgi?deptlookup=1&search=PUBH8142~001&term=1193
- Instructor Supplied Information Last Updated:
- 10 January 2014

## Spring 2016 | PUBH 8142 Section 001: Epidemiologic Uncertainty Analysis (54433)

- Instructor(s)
- Class Component:
- Lecture
- Credits:
- 2 Credits
- Grading Basis:
- S-N only
- Instructor Consent:
- No Special Consent Required
- Instruction Mode:
- In Person Term Based
- Times and Locations:
- Regular Academic Session01/19/2016 - 05/06/2016Tue 12:20PM - 02:15PMUMTC, East BankMayo Bldg/Additions 1155
- Also Offered:

- Course Catalog Description:
- Scientific interpretation of statistical analysis as dependent on both data and assumptions. Techniques that enable an investigator to incorporate uncertainty about assumptions into a quantitative analysis. prereq: 8140
- Class Description:
- An observed relative risk (RRobs) can be described mathematically as the product of the causal relative risk (RRcausal), a desired effect measure for etiologic epidemiologic studies, and error factors for the impact on study results of imperfections in the design, conduct and analysis of the study (uncontrolled confounding, losses-to-followup, nonrandom subject sampling, subject nonresponse, missing data, exposure and disease measurement error, unjustified statistical model assumptions, and random error). When viewed from this perspective, it becomes clear that RRcausal is not identifiable (i.e., cannot be validly estimated) without making assumptions about the values of the error-factor and random-error terms. A standard quantitative analysis does not account for most study imperfections. It therefore implicitly assumes that the product of the error factors equals 1.0. This standard-practice assumption, however, has neither theoretical nor empirical justification. We therefore advise epidemiologists to replace the standard assumption with more justifiable assumptions about the values of the error-factor terms. These more-justifiable assumptions can be incorporated into a quantitative analysis with uncertainty analysis (also known as bias modeling, probabilistic sensitivity analysis, Monte Carlo sensitivity analysis). We discuss this technique in this class.
- Textbooks:
- http://www.bookstores.umn.edu/buybooks.cgi?deptlookup=1&search=PUBH8142~001&term=1163
- Instructor Supplied Information Last Updated:
- 10 January 2014

## Spring 2014 | PUBH 8142 Section 001: Epidemiologic Uncertainty Analysis (60155)

- Instructor(s)
- Class Component:
- Lecture
- Credits:
- 2 Credits
- Grading Basis:
- S-N only
- Instructor Consent:
- No Special Consent Required
- Instruction Mode:
- In Person Term Based
- Class Attributes:
- Delivery Medium
- Times and Locations:
- Regular Academic Session01/21/2014 - 05/09/2014Tue 12:20PM - 02:15PMUMTC, East BankMayo Bldg/Additions 1155
- Also Offered:

- Course Catalog Description:
- Scientific interpretation of statistical analysis as dependent on both data and assumptions. Techniques that enable an investigator to incorporate uncertainty about assumptions into a quantitative analysis.
- Class Description:
- An observed relative risk (RRobs) can be described mathematically as the product of the causal relative risk (RRcausal), a desired effect measure for etiologic epidemiologic studies, and error factors for the impact on study results of imperfections in the design, conduct and analysis of the study (uncontrolled confounding, losses-to-followup, nonrandom subject sampling, subject nonresponse, missing data, exposure and disease measurement error, unjustified statistical model assumptions, and random error). When viewed from this perspective, it becomes clear that RRcausal is not identifiable (i.e., cannot be validly estimated) without making assumptions about the values of the error-factor and random-error terms. A standard quantitative analysis does not account for most study imperfections. It therefore implicitly assumes that the product of the error factors equals 1.0. This standard-practice assumption, however, has neither theoretical nor empirical justification. We therefore advise epidemiologists to replace the standard assumption with more justifiable assumptions about the values of the error-factor terms. These more-justifiable assumptions can be incorporated into a quantitative analysis with uncertainty analysis (also known as bias modeling, probabilistic sensitivity analysis, Monte Carlo sensitivity analysis). We discuss this technique in this class.
- Textbooks:
- http://www.bookstores.umn.edu/buybooks.cgi?deptlookup=1&search=PUBH8142~001&term=1143
- Instructor Supplied Information Last Updated:
- 10 January 2014

## Spring 2013 | PUBH 8142 Section 001: Epidemiologic Uncertainty Analysis (55582)

- Instructor(s)
- Class Component:
- Lecture
- Credits:
- 2 Credits
- Grading Basis:
- S-N only
- Instructor Consent:
- No Special Consent Required
- Instruction Mode:
- In Person Term Based
- Class Attributes:
- Delivery Medium
- Times and Locations:
- Regular Academic Session01/22/2013 - 05/10/2013Tue 12:20PM - 02:15PMUMTC, East BankMayo Bldg/Additions 1155
- Also Offered:

- Course Catalog Description:
- Scientific interpretation of statistical analysis as dependent on both data and assumptions. Techniques that enable an investigator to incorporate uncertainty about assumptions into a quantitative analysis.
- Class Description:
- An observed relative risk (RRobs) can be described mathematically as the product of the causal relative risk (RRcausal)?a desired effect measure for etiologic epidemiologic studies?and error factors for the impact on study results of imperfections in the design, conduct and analysis of the study (uncontrolled confounding, losses-to-followup, nonrandom subject sampling, subject nonresponse, missing data, exposure and disease measurement error, unjustified statistical model assumptions, and random error). When viewed from this perspective, it becomes clear that RRcausal is not identifiable (i.e., cannot be validly estimated) without making assumptions about the values of the error-factor and random-error terms. A standard quantitative analysis does not account for most study imperfections. It therefore implicitly assumes that the product of the error factors equals 1.0. This standard-practice assumption, however, has neither theoretical nor empirical justification. We therefore advise epidemiologists to replace the standard assumption with more justifiable assumptions about the values of the error-factor terms. These more-justifiable assumptions can be incorporated into a quantitative analysis with uncertainty analysis (also known as bias modeling, probabilistic sensitivity analysis, Monte Carlo sensitivity analysis). We discuss this technique in this class.
- Textbooks:
- http://www.bookstores.umn.edu/buybooks.cgi?deptlookup=1&search=PUBH8142~001&term=1133
- Instructor Supplied Information Last Updated:
- 6 November 2009

## ClassInfo Links - Public Health Classes

- To link directly to this ClassInfo page from your website or to save it as a bookmark, use:
- http://classinfo.umn.edu/?subject=PUBH&catalog_nbr=8142
- To see a URL-only list for use in the Faculty Center URL fields, use:
- http://classinfo.umn.edu/?subject=PUBH&catalog_nbr=8142&url=1
- To see this page output as XML, use:
- http://classinfo.umn.edu/?subject=PUBH&catalog_nbr=8142&xml=1
- To see this page output as JSON, use:
- http://classinfo.umn.edu/?subject=PUBH&catalog_nbr=8142&json=1
- To see this page output as CSV, use:
- http://classinfo.umn.edu/?subject=PUBH&catalog_nbr=8142&csv=1

ClassInfo created and maintained by the Humphrey School of Public Affairs.

If you have questions about specific courses, we strongly encourage you to contact the department where the course resides.