Measurement Error Webinar Series

The information on this page is archived and provided for reference purposes only.


Overview

Welcome to the Measurement Error Webinar Series, organized by collaborators from the National Cancer Institute, the NIH Office of Dietary Supplements, the United States Department of Agriculture, the Gertner Institute, Texas A&M University, and Wake Forest University. The series is intended for nutritionists, epidemiologists, statisticians, graduate students, and others with an interest in measurement error in dietary intake data. A basic level of familiarity with statistics and dietary assessment is recommended.

The goal of the webinar series is to provide participants with an understanding of:

  • the sources and magnitudes of dietary measurement errors;
  • how measurement error may affect estimates of usual dietary intake distributions;
  • how measurement error may affect analyses of diet-health relationships; and
  • how the effects of measurement error may be mitigated.

Concepts related to accounting for complex survey methods, estimating total intakes from diet and supplements, and the use of multiple dietary assessment instruments and self-report data along with biomarker data to reduce measurement error are also addressed.

An archive of the webinar series, which ran from September 20th to December 6th, 2011, is provided below. Session descriptions provide details on the topics covered and the objectives, recommended resources, and key terms for each webinar. In addition to the recordings, which can be viewed online, multiple versions of the slide sets are available for download. The slides with speaking notes will display two pages at a time by default to facilitate on-screen viewing. For economical printing, a version with six slides per page is also provided. Additional supporting materials, including a glossary, are available from the links at the right of this page.

Return to Top

Session Descriptions

Webinar 1

Introduction to the Problem of Measurement Error in Dietary Intake Data

Presenter: Sharon Kirkpatrick, National Cancer Institute


Session objectives
  • Define concepts related to usual dietary intake.
  • Identify random and systematic errors that may occur in dietary assessment and their impact on estimates of dietary intake.
  • Describe statistical concepts underpinning approaches to correcting for measurement error in self-report dietary intake data.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 2

Estimating Usual Intake Distributions for Dietary Components Consumed Daily by Nearly All Persons

Presenter: Kevin Dodd, National Cancer Institute


Session objectives
  • Identify considerations in estimating usual intakes of nutrients and foods consumed nearly daily by nearly all persons.
  • Describe statistical modeling techniques and associated data requirements for estimating usual intake distributions for nutrients and foods consumed nearly daily by nearly all persons.
  • Describe assumptions made in current approaches to estimating usual intake distributions and challenges in self-report dietary data not addressed by current statistical modeling techniques.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 3

Estimating Usual Intake Distributions for Dietary Components Consumed Episodically

Presenter: Janet Tooze, Wake Forest University


Session objectives
  • Define key concepts of food consumption related to usual intake estimation.
  • Identify challenges in estimating usual intake for episodically-consumed dietary constituents.
  • Explain statistical modeling approaches for estimating usual intake for episodically-consumed dietary constituents.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 4

Accounting for Complex Survey Design in Modeling Usual Intake

Presenter: Kevin Dodd, National Cancer Institute


Session objectives
  • Identify considerations in the analysis of dietary data collected as part of a complex survey, including stratification, clustering and weighting.
  • Identify methods of variance estimation for complex survey samples and how these are incorporated into estimation of usual intake distributions.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 5

Estimating Usual Total Nutrient Intake Distributions from Diet and Supplements

Presenter: Regan Bailey, Office of Dietary Supplements, National Institutes of Health


Session objectives
  • Identify key challenges and considerations in combining dietary and supplement intake data.
  • Explain statistical approaches to estimating total nutrient intakes.
  • Describe assumptions and caveats of current techniques of estimating total nutrient intakes.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 6

The Problem of Measurement Error When Examining Diet-Health Relationships

Presenter: Laurence Freedman, Gertner Institute


Session objectives
  • Explain the types and magnitude of measurement error that occur in dietary data.
  • Describe statistical models for evaluating diet-health relationships, including energy adjustment models.
  • Describe the qualitative and quantitative impact of measurement error on studies of diet-health relationships.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 7

Assessing Diet-Health Relationships with FFQ: Focus on Dietary Components Consumed Daily by Nearly All Persons

Presenter: Douglas Midthune, National Cancer Institute


Session objectives
  • Identify challenges in estimating diet-health relationships caused by measurement error in dietary assessment, with a focus on components consumed nearly daily by nearly all persons.
  • Describe approaches to correct estimated diet-health relationships for bias due to measurement error when diet is assessed by a food frequency questionnaire.
  • Understand the role of calibration studies in assessing and correcting for measurement error in dietary instruments.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 8

Assessing Diet-Health Relationships with FFQ: Focus on Episodically-Consumed Dietary Components

Presenter: Victor Kipnis, National Cancer Institute


Session objectives
  • Review statistical risk models for evaluating diet-health relationships in nutritional epidemiology.
  • Describe application of regression calibration to correct for FFQ measurement error using repeat short-term reference measurements in a substudy.
  • With focus on episodically-consumed dietary components, describe application of a new methodology to carry out regression calibration in risk models with energy-adjusted dietary covariates.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 9

Estimating usual intake distributions for multivariate dietary variables

Presenter: Raymond Carroll, Texas A&M University


Session objectives
  • Identify challenges in addressing measurement error when modeling multivariate dietary variables such as diet quality indices.
  • Describe statistical modeling techniques to correct for measurement error in estimating multivariate dietary variables.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 10

Combining Self-Report Dietary Assessment Instruments to Reduce the Effects of Measurement Error

Presenter: Douglas Midthune, National Cancer Institute


Session objectives
  • Describe methods of combining self-report dietary instruments to reduce measurement error and improve the power to detect diet-health associations.
  • Understand situations in which combining information from various instruments may provide the most benefit.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 11

Combining self-report dietary intake data and biomarker data to reduce the effects of measurement error

Presenter: Laurence Freedman, Gertner Institute


Session objectives
  • To describe the motivation for combining dietary self-reports and biomarkers.
  • To provide an overview of different methods of combining self-reports and biomarkers, their aims, and the knowledge required for implementing each method.
  • To identify the potential gains of such combinations and the limitations of the methods.
Recommended resources
Watch a recording of this session.External Web Site Policy

Webinar 12

Assessing diet-health relationships using a short-term unbiased dietary instrument: focus on risk models with multiple dietary components

Presenter: Victor Kipnis, National Cancer Institute


Session objectives
  • Identify challenges in estimating relationships between a dietary exposure measured by repeated application of a short-term unbiased instrument and a health outcome in a risk model with multiple dietary components.
  • Describe potential approaches to correct for within-person random measurement error in estimating relationships between a dietary exposure measured by a short-term unbiased instrument and a health outcome in a risk model with multiple dietary components.
Recommended resources
Watch a recording of this session.External Web Site Policy

Return to Top

Dedication


Photo of Dr. Arthur Schatzkin

This series is dedicated to the memory of Dr. Arthur Schatzkin.

In recognition of his internationally renowned contributions to the field of nutrition epidemiology and his commitment to understanding measurement error associated with dietary assessment.

Return to Top

The information on this page is archived and provided for reference purposes only.