Censored Data Analysis I - Basics

Location – (Web PDC)

Instructions regarding how to log onto the web PDC and the conference call phone number, as well as handouts, will emailed to the registered participants at least twenty four hours prior to the web PDC.

Instructor:

Dr. Paul Hewett Ph.D.

Course Description

This 2.5 hour webinar PDC consists of Part I of a two part introduction to Censored Data Analysis (CDA). CDA refers to the analysis and interpretation of datasets containing one or more occupational exposure measurements below the field limit of detection or quantitation (i.e., the Minimally Detectable Concentration or the Minimally Quantifiable Concentration). (See the list of Prerequisites.)

Part I covers the most commonly used CDA methods used to estimate the parameters of the exposure profile. These methods can be divided into families: substitution methods, log-probit regression methods, maximum likelihood estimation methods, and non-parametric methods. An estimate of the exposure profile 95th percentile or mean is then either calculated directly or calculated using the parameter estimates for the exposure profile. The various CDA methods vary in terms of bias, precision, overall accuracy, and ease of calculation. Recommendations will be presented for analyzing various types of censored datasets.

The webinar will last 2.5 hours, with an additional 30 minutes for followup questions and clarifications. Each participant will receive an email confirming participation in the PDC.

Learning Objectives

  • Recognition of the major issues and problems associated with the statistical analysis of censored datasets (i.e., datasets that contain one or more non-detects.
  • Familiarity with the four families of Censored Data Analysis (CDA) methods and most commonly used method within each family.
  • Practice analyzing and interpreting censored datasets using freeware software (IHDataAnalyst-LiteEdition).
  • Awareness of the bias and overall accuracy of the methods presented.
  • Recognition of the consequences of failing to properly analyze a censored dataset.

Prerequisites

  • Participation in the webinar PDC “IH Stats I – Basics” or a strong background in IH statistics.
  • Familiarity with the AIHA Exposure Assessment and Management Model (see the AIHA monograph “A Strategy for Assessing and Managing Occupational Exposures”, 3rd Edition (2006)
  • Working knowledge of the IHDataAnalyst (full or Student Edition).
  • Recommended reading prior to the WebPDC:
    • Hewett, P. and Ganser, G.H. (2007): A Comparison of Several Methods for Analyzing Censored Data. Annals of Occupational Hygiene 51:611-632.
    • “Appendix VIII – Analysis of Censored Data” in the AIHA monograph “in the AIHA monograph “A Strategy for Assessing and Managing Occupational Exposures”, 3rd Edition (2006).

Presentation List

  • Definitions, Assumptions, and Goals
  • Detects and Non-detects
  • Characteristics of Censored Datasets
  • Introduction to Parametric CDA Methods + Workshop
  • Introduction to Non-parametric CDA Methods + Workshop
  • Recommendations
  • Stump the Instructor – Analysis of Student Datasets

Fees:

$200 per participant

Cancelation Policy

A credit for an upcoming web PDC or a refund will be given for any cancellation made at least one business day prior to the day of the class. For example, if the web PDC is presented on a Saturday, the cancellation must be made by 6PM (1800) US Eastern time on the previous Thursday in order to qualify for a credit or refund. All refunds will be minus a $50 fee to recover credit card fees and administrative costs. No refunds or credits will be given if the notice of cancellation is not received within one business day of the web PDC.

Exposure Assessment Solutions, Inc. reserves the right to cancel the PDC at any time, in which case a full refund or credit will be given to each participant.

Read these notes before you sign up:

NOTE: The registration and participation policy for each Web PDC is the same as for a fixed location or on-site PDC. Each particpant must register. Registration is per participant, not per viewing site. Only those registered should view and particpant in this web PDC.

NOTE: The email address submitted at the time of payment will be used for all email communications and for logging onto the web PDC.

NOTE: This web PDC will presented using GoToMeeting. GoToMeeting must be installed on your computer. To install (usually takes less than two minutes), click on the GoToMeeting link above, scroll to the bottom of the page, and click on the link “Download GoToMeeting (for customers)”. If you do not have adminstrative privileges for your computer have your IT person install the GoToMeeting software well in advance of the web PDC.

NOTE: The conference call number will not be toll free. You will be responsible for any long distance phone charges if you do not have an unlimited long distance account. However, you may also connect using your computer speakers and microphone or headset (VoIP).

NOTE: See the “Prerequisites” below for software and freeware that will be helpful when taking this web PDC. To download and install the program(s) you must first be a registered member of www.oesh.com.