|Title:||Hierarchical Network Models for Education Research|
|Principal Investigator:||Junker, Brian||Awardee:||Carnegie Mellon University|
|Program:||Statistical and Research Methodology in Education [Program Details]|
|Award Period:||3 years (7/1/12–6/30/15)||Award Amount:||$480,986|
|Type:||Methodological Innovation||Award Number:||R305D120004|
Experimental and observational studies in education are sometimes focused not on the effects of changing curriculum, teaching and learning materials, or classroom technique, but rather on changes in the way students, teachers, teaching coaches, and administrators work with one another. In short, many studies focus on changes in student and professional social networks in school systems. Social network data are often ignored in education research, even though social networks play a role in education processes and can both influence the function of an intervention and be influenced by an intervention in an educational research setting. Social network analysis (SNA) is a collection of quantitative methods for comparing and measuring relationships among individuals in a network. SNA has been used as the basis of analyses of interpersonal relationships in clubs and other social groups; analysis of academic paper co-authorship and citations; the development of online commercial recommender systems; and much more.
In this project, researchers will carry out a methodological investigation of social network analysis and examine its potential applicability in education through the development of hierarchical network models (HNMs). HNMs can be used to model multiple, partially-exchangeable social networks (such as the professional communities that teachers form in multiple different school buildings), incorporate treatment and other effects, and accommodate the usual nesting/cluster structure that hierarchical linear models handle in other contexts. In addition to conducting simulation studies of issues pertinent to statistical power and model estimation, the researchers will use hierarchical network models to analyze available social network data, including data from the National Longitudinal Study of Adolescent Health. Products of the research will be peer-reviewed publications and user-friendly open-source code for conducting analyses with HNM.
Project Website: http://hnm.stat.cmu.edu/
Sweet, T. M. (2017). Modeling Collaboration with Social Network Models. In von Davier A., Zhu M., Kyllonen P. (Eds) Innovative Assessment of Collaboration (pp. 287–302). Springer, Cham.
Sweet, T.M., Thomas, A.C., and Junker, B.W. (2014). Hierarchical Mixed Membership Stochastic Block Models for Multiple Networks and Experimental Interventions. Handbook on Mixed Membership Models and Their Applications. CRC Press: Taylor and Francis Group.
Journal article, monograph, or newsletter
Hopkins, M., Lowenhaupt, R., and Sweet, T.M. (2015). Organizing English Learner Instruction in New Immigrant Destinations: District Infrastructure and Subject-Specific School Practice. American Educational Research Journal, 52(3): 408–439.
Spillane, J. P., Hopkins, M., and Sweet, T. (2015). Intra- and Interschool Interactions about Instruction: Exploring the Conditions for Social Capital Development. American Journal of Education, 122(1): 71–110.
Sweet, T. M. (2015). Incorporating Covariates into Stochastic Blockmodels. Journal of Educational and Behavioral Statistics, 40(6): 635–664.
Sweet, T. M., Thomas, A. C., and Junker, B. W. (2013). Hierarchical Network Models for Education Research: Hierarchical Latent Space Models. Journal of Educational and Behavioral Research, 38(3): 295–318.
Sweet, T.M., and Junker, B.W. (2016). Power to Detect Intervention Effects on Ensembles of Social Networks. Journal of Educational and Behavioral Statistics, 41(2): 180–204.
Sweet, T.M., and Zheng, Q. (2017). A Mixed Membership Model-Based Measure for Subgroup Integration in Social Networks. Social Networks, 48, 169–180.