\documentclass[11pt]{article}
\setlength{\oddsidemargin}{0.0truein}
\setlength{\evensidemargin}{0.0truein}
\setlength{\textwidth}{6.5truein}
\setlength{\topmargin}{0.0truein}
\setlength{\textheight}{9.0truein} \setlength{\headsep}{0.0truein}
\setlength{\headheight}{0.0truein} \setlength{\topskip}{10.0pt}
\setlength{\parskip}{5mm}
\usepackage{url}
\begin{document}

\begin{center}
  \textbf{\textsc{STANFORD UNIVERSITY}}\\[5pt]
  \textbf{\textsc{DEPARTMENT OF STATISTICS}}\\[5pt]
  \Large{\textbf\textsc{{DEPARTMENTAL SEMINAR}}}
\end{center}

\begin{center}
  4:15 p.m., Tuesday, May 16, 2006\\
  Sequoia Hall Room 200\\
  (Cookies at 3:45 in 1st Floor Lounge)
\end{center}

\begin{center}
  \textsl{ Raja Velu   }\\
Whitman School of Management, Syracuse University
\end{center}

\begin{center}
  \textbf{ Reduced-Rank Regression Models for Longitudinal Data    } 
\end{center}

\noindent
The random effects models discussed in Laird and Ware (1982) are widely
used in the analysis of longitudinal data. The principal advantage of the 
models is that they provide a unified framework to deal with a variety of 
important special problems that arise in practice. The models can account 
for serial correlations and also deal with the unbalanced data. The 
two-stage modeling approach allow for modeling both between and within 
unit variations. But when the dimensions of the response and the predictor 
variables are large, it is possible that there is some redundancy in the 
number of parameters. We will demonstrate how reduced-rank regression 
methods can be used to build more parsimonious models. The ideas can also 
be related to latent variable models. An example based on scanner data will 
be presented to illustrate the methods. 
 
\noindent
Raja Velu is a Professor, Whitman School of Management at Syracuse University.
Currently he is on sabbatical visiting the statistics department at Stanford 
and at IBM-Almaden. 

\end{document}