Syllabus: SOC WL 590 – Fall 2013

SEMINAR IN LONGITUDINAL DATA ANALYSIS

Instructor:

Elena Erosheva  

C 14C, Padelford Hall (CSSS)

elena at stat.washington.edu
  • Class times: Tuesday and Thursday 10:30 - 11:50
  • Place: Parrington Hall 306
  • Office hours:  Monday 2:30-3:20, Wednesday 2:00-3:00, or by appointment 
  • Web: follow the class link from my homepage at http://www.stat.washington.edu/elena
  • Questions by e-mail are welcome. They will often be answered quite quickly, but this is not guaranteed. In particularly, I don't always check e-mail over weekends.

Seminar description

Understanding changes, including those that result from targeted interventions, is central to much empirical research in the social and behavioral sciences. The goal of this seminar is to explore how research questions about change (and event occurrences) can be addressed with longitudinal data analysis. During the first half of the seminar, we will follow Part I of Singer and Willett’s Applied Longitudinal Data Analysis to learn why study change, what are the minimum data requirements for investigating change over time, what techniques are available for exploratory analysis specific to longitudinal data, and what statistical models are appropriate. We will focus on fundamental models for longitudinal analysis that have been variously known as individual growth models, random coefficient models, growth curve models, multilevel models, mixed effects models and hierarchical linear models.

 

During the second part of the course, we will discuss longitudinal data analysis topics that will be selected by the seminar participants.  These include but are not limited to:

·         Modeling discontinuous and nonlinear change over time.

·         Reconciling terminology across mixed effects, hierarchical and multilevel modeling traditions. 

·         Mixture models for life course data (i.e., variations on growth mixture and group-based trajectory models).

·         Assessing effectiveness of interventions.

·         Dynamic treatment regimes.

·         Latent variable models for change.

·         Age, period, and cohort effects.

·         Accelerated longitudinal designs.

·         Power analysis for longitudinal data.

·         Dealing with missing data in longitudinal analysis.

 

 

Course Text

 Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence (2003) Singer, J.D. and Willett, J. B.

Course objectives

·         To investigate how research questions about change can be addressed with longitudinal data. 

·         To study multilevel, growth curve and hierarchical modeling in application to longitudinal data.

·         To provide an open forum for scholarly exchange on the longitudinal data analysis issues of interest to the class participants.

Computing

The textbook  has a companion web site that presents data sets and computer code in a wide variety of software packages that include R, Stata, SPSS and Mplus:

http://www.ats.ucla.edu/stat/examples/alda.htm

Participants will be free to use a package of their choice for presentations (literature review presentations are not required to include a computing code component).

Prerequisites

SOC 504-505-506 or equivalent. Solid knowledge of linear regression.

Homework assignments and grades

  • The course will be graded credit/no credit.
  • Credit will be given for attendance, participation, and a research presentation. 
  • Attendance: Students are expected to attend all lectures and research presentations. Contact the instructor in advance if you expect to be absent because of conference travel or other schedule conflicts.
  • Participation: Students are expected to stay on top of the reading assignments and participate during in-class discussions. The class will be in a seminar format rather than a lecture style, with the goal to stimulate discussions among class participants.
  • Research presentation: Each student will choose a research topic that must be related to longitudinal data analysis in a broad sense. For example, this could be a review of one of the longitudinal data analysis topics listed above, a review of longitudinal data analysis methods in a certain field (e.g., mental health), or a hands-on implementation and demonstration of certain longitudinal data analysis techniques. In consultation with the instructor, students are expected to find material for exploring their research topic, to prepare a research presentation for an in-class delivery during the quarter, to recommend readings on the topic for other class participants, and to submit written materials based on the presentation. Depending on the project choice, written materials can be in the form of a summary handout, a brief report summarizing the presentation, a review paper, or, in case of a didactic data analysis, a collection of slides. Written materials will be due one week after the scheduled presentation.
  • Joint presentations: In an unlikely event that the final enrollment will have exceeded slots available for presentations, some research topics may be shared by pairs on individuals. If you and your partner have a plan for splitting the work on a research topic that you both agree on, please contact the instructor. 

Students with Disabilities

If you would like to request academic accommodations due to a disability, please contact Disabled Student Services, 448 Schmitz, 543-8924 (V/TTY).  If you have a letter from Disabled Student Services indicating you have a disability that requires academic accommodations, please present the letter to me so we can discuss the accommodations you might need for this class.