Judith D. Singer

Judith D. Singer (Ph.D., Statistics, Harvard University) is the James Bryant Conant Professor of Education and former Academic Dean at the Harvard Graduate School of Education. An internationally renowned social statistician, her professional life focuses on improving the quantitative methods used in social, educational and behavioral research. Singer is primarily known for her contributions to the practice of multilevel modeling, survival analysis, and individual growth modeling, and to making these and other statistical methods accessible to empirical researchers.

Singer's wide-ranging interests have led her to publish across a broad array of disciplines including: statistics, education, psychology, and medicine/public health. In addition to writing (or more often co-writing) nearly 100 papers and book chapters, she has also co-written three books including By Design: Planning Better Research in Higher Education and Who Will Teach: Policies that Matter, both published by Harvard University Press. Here is her curriculum vitae.

Her most recent book with longtime collaborator John B. Willett is Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence (NY: Oxford University Press), for which they received Honorable Mention from the American Publishers Association for the best Mathematics & Statistics book of 2003. ALDA offers an accessible in-depth presentation of two popular statistical methods for analyzing longitudinal data: multilevel modeling of individual change and hazard/survival modeling for event occurrence (in both discrete- and continuous-time). The HGSE ALDA website includes loads of useful materials. To download the ALDA data and free copies of all the computer programs necessary to reproduce the analyses in a variety of software packages (including SAS, stata, SPSS, MLwiN and HLM) visit the companion UCLA ALDA website. For a limited time, Oxford is generously offering a 20% discount coupon on ALDA (bringing the price down to $55.50).

Willett and Singer have recently begun their next collaborative venture, writing a new book on multilevel modeling tentatively titled Applied Multilevel Data Analysis (AMDA). Written with the same tone and emphasis as ALDA, AMDA will offer an accessible in-depth presentation of the many approaches for modeling multilevel data structures, integrating ideas from multiple disciplines including statistics, econometrics and survey sampling. Check back periodically for publication updates.

Singer has received numerous awards for her work, including a fellowship at the Center for Advanced Study in the Behavioral Sciences and election to the National Academy of Education. Along with her collaborators, the American Educational Research Association (AERA) has given her the Raymond B. Cattell Award, the Review of Research Award, and the Palmer O. Johnson Award.

COURSES

DOWNLOADABLE PAPERS

• Using SAS PROC MIXED to fit multilevel models, hierarchical models, and individual growth models, Journal of Educational and Behavioral Statistics. It is written as a step-by-step tutorial that shows how to use SAS to fit the two most common multilevel models: (1) two-level models, designed for data on individuals nested within naturally occuring hierarchies (e.g., students within classes) and (2) individual growth models, designed for exploring longitudinal data (on individuals) over time. The conclusion provides code for three level models and an appendix provides code for working with multilevel data in SAS.
• Great linked resource: The UCLA Academic Computing Center has created a wonderful companion website for the SAS PROC MIXED paper that includes the actual data sets and computer code for replicating all the analyses in that paper in stata xtmixed, SPSS mixed, SPLUS lme, HLM and MLwiN.
• The design and analysis of longitudinal studies of development and psychopathology in context: Statistical models and methodological recommendations, Development and Psychopathology. It provides a concise overview of how researchers can use both individual growth modeling and the methods of survival analysis to explore development over time. In addition to presenting the fundamental statistical models, the paper presents recommendations for research design and measurement.
• Early child care selection: Variation by geographic location, maternal characteristics, and family structure, Developmental Psychology. Analyzing data from a national probability sample of children under 6, we use discrete-time survival analysis to examine the predictors of age at first placement into formal day care.
• Maternal correlates of growth in toddler vocabulary production in low-income families. Child Development. Uses non-linear individual growth models to examine maternal correlates of growth in toddlers' vocabulary production collected three times from from ages 1 to 3.
• Afraid to discuss evolution?, Chance. This is 3-page op-ed piece that exposes the flawed statistical reasoning presented in a February 2005 New York Times editorial presenting a misleading statistic that 41% of the biology teachers in Lousiana reject evolution.
• Lessons from the Red Sox Playbook, Harvard Education Letter. This is 2-page op-ed piece suggesting that education professionals have much to learn by adopting the kind of rich quantitative analysis--and an orientation towards experimentation--that helped the Boston Red Sox finally "reverse the curse" and win the World Series in 2004.

You can download copies of many other papers I've co-written with John Willett by going to his webpages for papers on survival analysis, research design, the teaching of applied statistics, and general statistical concerns.

RECENT PRESENTATIONS

Questions to Ask When Reporting Education Research--Invited Workshop for the 3rd Annual Education Research and Statistics Bootcamp of the Education Wrtiers Association and the Nieman Foundation for Journalism, Cambridge, MA, February 2007.

Longitudinal Research: Current Status and Future Prospects -- Invited Keynote Address at the 45th Congress of the German Psychological Association, Nurnberg, Germany, September 2006

Invited Short Course: Individual Growth Modeling: Modern Methods for Studying Change, June 2006, Munich, Germany.

• Course handout
• Background reading for all participants--Chapters 1 and 2 of ALDA
• Overview of the Short Course
• Chapter 3: Introducing the Multilevel Model for Change
• Chapter 4: Doing Data Analysis with the Multilevel Model for Change
• Chapter 5: Treating TIME More Flexibly
• Chapter 6: Modeling Discontinuous and Nonlinear Change
• Using SAS Proc Mixed to fit the multilevel model for change

Longitudinal Research: Current Status and Future Prospects -- Kendon Smith Lecture, University of North Carolina, Greensboro, November 2005. Also given as a 2005 Back to School Master Lecture at the Biennial Meeting of the Society for Research on Child Development (SRCD)

Thoughts on Software for Multilevel Modeling -- Invited Commentary at the 2005 Annual meeting of American Educational Research Association (AERA)

Invited Workshop at the Lister Hill Center, University of Alabama Birmingham, School of Public Health , April 2005

• Longitudinal Research: Current Status and Future Prospects
• Discrete Time Survival Analysis
• Continuous Time Survival Analysis

OLDER POWERPOINT PRESENTATIONS FOR ALDA WORKSHOPS

• Overview of the Workshop
• Chapter 3: Introducing the Multilevel Model for Change
• Chapter 4: Doing Data Analysis with the Multilevel Model for Change
• Chapter 5: Treating TIME More Flexibly
• Chapter 6: Modeling Discontinuous and Nonlinear Change
• Chapters 10 and 11: Describing Discrete-Time Event Occurrence Data and Fitting Basic Discrete-Time Hazard Models
• Chapter 12: Extending the Discrete-Time Hazard Model
• Chapter 13: Describing Continuous-Time Event Occurrence Data
• Chapters 14 and 15: Fitting Cox Regression Models

LINKS of POTENTIAL INTEREST

Statistical sites

• UCLA Multilevel modeling portal
• Searchable archive of multilevel modeling newsgroup

Statistical software

• Statistical Computing Resources at UCLA
• SAS
• STATA
• SUDAAN
• WESVAR-PC

Archives

• Harvard Government Data Center
• Inter-university Consortium for Political and Social Research
• National Center for Education Statistics
• US Census Bureau
• Bureau of Labor Statistics
• National Center for Health Statistics
• National Data Archive on Child Abuse and Neglect (at Cornell University)

Selected statistics departments with interesting resources

• Harvard Statistics Dept
• Harvard Biostatistics Dept
• Arizona State University, Research Methods Resources from the College of Education
• Statlib at Carnegie Mellon
• Virtual statistics library at the Univ of Florida
• Virtual biostatistics library at the Univ of Washington

Resources for teaching statistics

• CHANCE course (using newspaper articles to teach statistics)
• Journal of Statistics Education
• Data and story library
• Data surfing on the World Wide Web
• An incredibly cool collection of java applets for learning statistical concepts. Check it out!

Professional organizations

• American Educational Research Association
• American Statistical Association
• American Psychological Association
• American Psychological Society
• American Public Health Association

---

Last Modified May 2008
HGSE Disclaimers and Publishing Policies

These pages are copyrighted, © Judith D. Singer, 2008