Ph.D., Yale University
Professor Mirowsky studies social aspects of health and well being, particularly as they evolve over the life course. His current research looks at the effects on health and well being of education, employment, economic hardship, age at first birth, neighborhood disadvantage and disorder, and the sense of personal control. His recent or forthcoming publications appear in the Journal of Health and Social Behavior, Social Psychology Quarterly, the American Sociological Review, Social Forces, the Journal of Marriage and Family, Social Science and Medicine, Demography, Milbank Quarterly, Sociological Methods and Research, the Journal of Gerontology: Social Sciences, and City and Community. The Institute for Scientific Information lists him as a highly cited social science researcher. He is currently a principal investigator, with Catherine E. Ross, on grants to study aging, status and the sense of control (NIA R01 AG12393), and education, resource substitution and health (NIA R01-AG023380).
John Mirowsky is a past editor of the Journal of Health and Social Behavior. He is a winner of the Leonard I. Pearlin award for lifetime achievement in the sociology of mental health. He also is co-winner, along with Catherine E. Ross, of the Leo G. Reeder Award for lifetime achievement in medical sociology.
John Mirowsky received his Ph.D. from Yale University in 1981, where he had a U.S. Public Health Service fellowship in the Sociology Department's Psychiatric Epidemiology Training Program. He spent his early career teaching epidemiology and biostatistics to medical students at the University of Illinois College of Medicine in Champaign Urbana. In 1984 he was tenured in the Division of Medical Information Sciences of the College of Medicine and the Department of Sociology of the College of Liberal Arts and Sciences at Illinois. He has been a professor of sociology at the University of Illinois, the Ohio State University, and the University of Texas.
SOC 395J • Educ, Socioecon Stat, Hlth
45785 • Spring 2012
Meets TH 1200pm-300pm MAI 1704
People with higher socioeconomic status have better health than lower status individuals. This class explores the links between social status and health, with a focus on education. Education forms a unique dimension of social status with qualities that make it especially important to health. It influences health in ways that are varied, present at all stages of adult life, cumulative, self-amplifying, and uniformly positive. Education creates desirable outcomes because it trains individuals to acquire, evaluate, and use information. It teaches individuals to tap the power of knowledge. Education develops the learned effectiveness that enables self-direction toward any and all values sought, including health.
SOC 395J • Struc Eq Modl Of Hlth/Life Crs
45780 • Fall 2010
Meets W 1200pm-300pm MAI 1704
This class is a practicum on structural equation modeling (SEM) of health and the life course. A structural equation model integrates two or more regression equations, showing the implications of the relationships among them. SEM developed from two older methods called path analysis and factor analysis. A path analysis shows the mediating links between variables. The classic path analyses in sociology showed how parental social status influenced a worker's educational attainment, which in turn influenced the worker’s occupational status, which in turn influenced income. A factor analysis infers the unseen common causes implied by the correlations among variables. When combined in SEM, these methods allow researchers to model complex sets of causal relationships.
SOC 395J • Struc Eq Modl Of Hlth/Life Crs
46785 • Fall 2009
Meets W 1200-300pm MAI 1704
Syllabus for Mirowsky’s SOC395J, UTx Fall 2009 (printed 8/23/09) Page 1
Structural Equation Models of Health and the Life Course
SOC 395J, Unique Number 46785, U. Texas, Austin, Fall 2009
Professor John Mirowsky
Class Time and Place
Wednesday 12:00 - 3:00 pm
This class is a practicum on structural equation modeling (SEM) of health and the life course. A
structural equation model integrates two or more regression equations, showing the implications
of the relationships among them. SEM developed from two older methods called path analysis
and factor analysis. A path analysis shows the mediating links between variables. The classic
path analyses in sociology showed how parental social status influenced a worker’s educational
attainment, which in turn influenced the worker’s occupational status, which in turn influenced
income. A factor analysis infers the unseen common causes implied by the correlations among
variables. When combined in SEM, these methods allow researchers to model complex sets of
All students must know multiple regression and have experience estimating and interpreting
regression equations. Structural Equation Models are sets of interrelated regression equations.
Most lectures and demonstrations will use the AMOS Graphics SEM program. AMOS allows
users to define the model by drawing a schematic called a path diagram. This makes it relatively
easy to learn the concepts and procedures. Other programs generally require instructions, similar
to the syntax of SAS, STATA or SPSS procedures, that take weeks or months to learn. AMOS is
available through the UT STATS server.
Class members can download the free student version of AMOS and the Acrobat file “Get
Running with AMOS Graphics” at the following site:
Class members also should read the introductory materials at the University of Texas ITS
Research Consulting site.
Syllabus for Mirowsky’s SOC395J, UTx Fall 2009 (printed 8/23/09) Page 2
Pay special attention to Wheaton’s alienation models in the examples. They represent the classic
sociological application of SEM.
Several other SEM programs are widely used, particularly LISREL (the first one), EQS, Mx and
Mplus. AMOS is easier to learn and to use than the others, but currently is not as powerful.
Mplus and Mx are the most powerful. Mplus is designed for sociological applications and Mx
for biological and psychological applications. Mx is free and can be downloaded at the following
Any student who wants to learn or use one of the other programs may do so, but in-class help
with problems may be limited. Groups of students who want to learn or use one of the other
programs are encouraged to work together.
Once the course gets going, most classes will have two parts. The first part will introduce formal
elements of SEM such as definitions, diagraming conventions, and technical issues such as
specification, identification, start values, convergence and model fit. These will be blackboard or
live computer demonstrations, usually by the instructor. The second part will be work on
common or individual projects.
Grades will be based on participation and term projects. In early sessions we will discuss a
variety of options. Some students may have well-developed ongoing projects related to thesis or
dissertation work that they want to develop. Others may want to work in a small group to
develop a common project that might eventually become a research paper submitted to a
conference or journal. We will work out expectations appropriate to each student’s level and
Readings will be posted on Blackboard as files or links to articles and web pages.