Dealing With Missing Data

About this course: This course will cover the steps used in weighting sample surveys, including methods for adjusting for nonresponse and using data external to the survey for calibration. Among the techniques discussed are adjustments using estimated response propensities, poststratification, raking, and general regression estimation. Alternative techniques for imputing values for missing items will be discussed. For both weighting and imputation, the capabilities of different statistical software packages will be covered, including R®, Stata®, and SAS®.

Who is this class for: This course is aimed at undergraduates, graduate students, and working professionals who have an interest and need in preparing survey data for analysis and distribution to data users.

Created by:  University of Maryland, College Park

• Taught by:  Richard Valliant, Ph.D., Research Professor

  Joint Program in Survey Methodology

Basic Info Course 5 of 7 in the Survey Data Collection and Analytics Specialization Commitment 4 weeks of study, 1-2 hours/week Language English How To Pass Pass all graded assignments to complete the course. User Ratings 3.8 stars Average User Rating 3.8See what learners said Coursework

Each course is like an interactive textbook, featuring pre-recorded videos, quizzes and projects.

Help from your peers

Connect with thousands of other learners and debate ideas, discuss course material, and get help mastering concepts.

Certificates

Earn official recognition for your work, and share your success with friends, colleagues, and employers.

University of Maryland, College Park The University of Maryland is the state's flagship university and one of the nation's preeminent public research universities. A global leader in research, entrepreneurship and innovation, the university is home to more than 37,000 students, 9,000 faculty and staff, and 250 academic programs. Its faculty includes three Nobel laureates, three Pulitzer Prize winners, 47 members of the national academies and scores of Fulbright scholars. The institution has a $1.8 billion operating budget, secures $500 million annually in external research funding and recently completed a $1 billion fundraising campaign.

Syllabus


WEEK 1


General Steps in Weighting



Weights are used to expand a sample to a population. To accomplish this, the weights may correct for coverage errors in the sampling frame, adjust for nonresponse, and reduce variances of estimators by incorporating covariates. The series of steps needed to do this are covered in Module 1.


7 videos, 7 readings expand

1. Video: Introduction
2. Reading: Class notes + additional reading
3. Video: Quantities to Estimate
4. Reading: Class notes
5. Video: Goals of Estimation
6. Reading: Class Notes
7. Video: Statistical Interpretation of Estimates
8. Reading: Class Notes
9. Video: Coverage Problems
10. Reading: Class Notes
11. Video: Improving Precision
12. Reading: Class Notes
13. Video: Effects of Weighting on SEs
14. Reading: Class Notes

Graded: Introductory quiz on weights
Graded: Quantities
Graded: Goals
Graded: Interpretation
Graded: Coverage
Graded: Improving precision
Graded: Effects on SEs

WEEK 2


Specific Steps



Specific steps in weighting include computing base weights, adjusting if there are cases whose eligibility we are unsure of, adjusting for nonresponse, and using covariates to calibrate the sample to external population controls. We flesh out the general steps with specific details here.


6 videos, 6 readings expand

1. Video: Overview
2. Reading: Class Notes
3. Video: Base Weights
4. Reading: Class Notes
5. Video: Nonresponse Adjustments
6. Reading: Class Notes
7. Video: Response Propensities
8. Reading: Class Notes
9. Video: Tree algorithms
10. Reading: Class Notes
11. Video: Calibration
12. Reading: Class Notes

Graded: Overview
Graded: Base weights
Graded: Nonresponse
Graded: Trees
Graded: Calibration

WEEK 3


Implementing the Steps
Software is critical to implementing the steps, but the R system is an excellent source of free routines. This module covers several R packages, including sampling, survey, and PracTools that will select samples and compute weights.


6 videos, 5 readings, 3 practice quizzes expand

1. Video: Software
2. Reading: Class Notes
3. Video: Base Weights
4. Reading: Class Notes + Software
5. Video: on Base Weights
6. Reading: Class Notes
7. Practice Quiz: Quiz on base weights
8. Video: Nonresponse Adjustments
9. Reading: Class Notes + Software for propensity classes
10. Practice Quiz: Quiz on nonresponse adjustments
11. Video: Examples of Calibration
12. Video: Software for Poststratification
13. Reading: Class Notes + Software for calibration
14. Practice Quiz: Quiz on calibration and poststratification

Graded: Software

WEEK 4


Imputing for Missing Items



In most surveys there will be items for which respondents do not provide information, even though the respondent completed enough of the data collection instrument to be considered "complete". If only the cases with all items present are retained when fitting a model, quite a few cases may be excluded from the analysis. Imputing for the missing items avoids dropping the missing cases. We cover methods of doing the imputing and of reflecting the effects of imputations on standard errors in this module.


6 videos, 5 readings expand

1. Video: Reasons for Imputation
2. Reading: Class Notes
3. Video: Means and hotdeck
4. Reading: Class Notes
5. Video: Regression Imputation
6. Reading: Class Notes
7. Video: Effect on Variances
8. Reading: Class Notes
9. Video: mice R package
10. Video: mice example
11. Reading: Class Notes + mice R package

Graded: Reasons for imputing
Graded: Means and hot deck
Graded: Regression imputation
Graded: Effects on variances
Graded: Imputation software

Summary of Course 5
We briefly summarize the methods of weighting and imputation that were covered in Course 5.


1 video, 1 reading expand

1. Video: Summary
2. Reading: Class Notes