Kenneth A. Bollen
Ph.D., Henry Rudolph Immerwahr Professor,
Psychology & Sociology
Campus Office: Davie Hall, Room 304
Campus Phone Number: (919) 962-7501
Dr. Bollen's Curriculum Vitae
Dr. Bollen's Personal Home Page
Dr. Bollen's Google Scholar profile
Dr. Bollen's publications in PubMed
Dr. Bollen's CPC publications
Bollen has advanced the modeling of latent variables in several ways, such as latent growth curve modeling, the development of nontraditional measurement theory with causal indicators, new robust estimators for latent variable models including instrumental variable techniques, furthering the understanding of causality in structural equation models, and proposing new approximations to Bayes Factors to help in selecting between competing models. The new test statistics, estimators, and models that he has created have been applied in population research to improve measurement, estimation, and assessing model fit. For instance, Bollen and colleagues developed and estimated a latent variable model that takes account of multiple indicators and measurement error in an analysis of fertility in Ghana and Peru. This study, published in Population Studies, draws attention to the distinction between “causal indicators” and “effect indicators” in demographic research. In Bollen's collaboration with CPC fellow Linda Adair, they have analyzed the developmental origins of adult disease risk factors with longitudinal data from the Cebu Longitudinal Health and Nutrition Survey, which has followed a cohort of births into adulthood. A Statistics in Medicine article from this project analyzes fetal conditions as a latent variable that mediates the effects of the mother's traits (e.g., smoking, age, parity) on birth weight, birth length, and gestational age. In another project related publication, he and coauthors analyze the error in blood pressure measures from the Cebu data over time using confirmatory factor analysis with controls for systematic error. Furthermore, Bollen provides statistical advice to many CPC Fellows and students.
Bollen will continue his work with Adair that applies latent growth curve models to analyze the levels and rates of growth in height and weight and their health consequences on the respondents using the Cebu data. He also will work on a similar "fetal origins" project for the United States as Subproject III on the new round of the Add Health NIH grant. In addition, Bollen is collaborating with RTI researchers in investigating diagnostic tests to determine whether weights are required when analyzing complex survey data. Furthermore, Bollen is continuing work on the development of new and very general longitudinal models that researchers can apply to population panel data. As a CPC preceptor he also is working with a CPC trainee investigating the connections between graphical models and structural equation models.
Primary Research Areas:
Current CPC Grants and Contracts:
Information updated on 5/21/2015