Network Science & Statistics

Networks are complex data objects that routinely violate the assumptions of more traditional statistical methods. New methods for analyzing network data are being developed all the time. My focus is on methods for analyzing networks whose individual elements, nodes and edges, do not follow the same data generating process: i.e., heterogeneous networks. By improving our understanding of network properties and how best to analyze them, I hope to support the use of network methods in psychology, neuroscience, and beyond. 

Representative Work

Henry, T. R., Ye, A., (In Prep) The Issue of Endogeneity in Network Psychometric Modelling.

Henry, T. R., Fisher, Z., & Bollen, K. (In Review). Bayesian model selection for model implied instrumental variable models.  

Henry, T. R., Gates, K., Prinstein, M., & Steinley, D. (2019). Modeling heterogeneous peer assortment effects using finite mixture exponential random graph models. Psychometrika.

Henry, T. R., Gesell, S., & Ip, E. (2016). Social position predicting physical activity level in youth: An application of hidden Markov modeling on network statistics. Social Computing, Behavioral-Cultural Modeling and Prediction (Conference Proceedings 2016 Volume).

Henry, T. R., Gesell, S., & Ip, E. (2016). Analyzing heterogeneity in the effects of physical activity in children on social network structure and peer selection dynamics. Network Science, 4, 336–363. doi: 10.1017/nws.2016.2


Functional Neuroimaging

One of the more recent focuses in neuroimaging is on functional connectivity, or how different regions of the brain activate together. My interest in this area is both methodological and substantive. In terms of methods research, I'm want to develop new methods that improve functional connectivity and the use of networks in neuroimaging, including finding better ways to examine between-subject heterogeneity. In terms of substantive research, I'm interested in developmental disorders (e.g., autism spectrum disorder, ADHD) and clinical disorders such as schizophrenia.

Representative Work

Henry, T. R., Duffy, K., Rudolph, M. D., Mostofsky, S. H. & Cohen, J. R. (2019) Bridging global and local topology in whole brain networks using the network statistic jackknife. Network Neuroscience.

Henry, T. R., Feczko, E., Cordova, M., Earl, E., Williams, S., Nigg, J. T., Fair D. A., & Gates, K. (2019). Comparing functional connectivity between groups with confirmatory subgrouping GIMME. NeuroImage.

Henry, T. R., & Cohen., J. (2019). Dysfunctional brain network organization in neurodevelopmental disorders. In P. J. Laurienti, B. Munsell, & G. Wu. (Eds.) Connectomics: Methods, Mathematical Models and Applications.

Henry, T. R., Dichter, G., & Gates, K.. (2018). Age and gender effects on intrinsic connectivity in autism using functional integration and segregation. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging.

Henry, T. R., & Gates, K. (2016). Causal search procedures for fMRI: Review and suggestions. Behaviormetrika. doi: 10.1007/s41237-016-0010-8


Natural Language Processing

Natural language processing is the use of statistics to obtain information from textual data. I am interested in the intersection of network science and natural language processing, such as the properties of networks of text like Wikipedia, or blog networks. Additionally, I am interested in topic modeling, or grouping text data into intelligible topics, a technique that has a number of similarities to modelling of heterogeneous networks.

Representative Work

Jackson, J. C., Watts, J., Henry, T. R., List, J., Forkel, R., Greenhill, S., Gray, R., &  Lindquist, K. (In Revision) Emotion varies in semantic structure across language families. Science. 

Henry, T. R., Banks D., Chai C., Owens-Oas, D. (2019). Modeling community structure and  topics in dynamic text  networks. Journal of Classification.