Futoshi Yumoto

Dr. Yumoto earned his PhD in Measurement, Statistics and Evaluation from the University of Maryland, College Park (2011). He is currently the Executive Director of Data Science at Resonate Networks, Inc., and served as Director of Data Science in Enterprise Business Intelligence at Comcast from 2016-2017. Dr. Yumoto has been a frequent collaborator with Dr. Tractenberg since 2006. He has over 17 years’ experience as a psychometrician and data scientist, employing techniques such as item response theory, mixture models, and Bayesian modeling/Bayesian Network building in applications ranging from state-wide achievement testing in K-12 contexts to Fortune 500 corporations in insurance, banking and retail industries, including public and private sector work as well as university research. Dr. Yumoto brings expert-level proficiency with techniques such as item response theory, mixture models, structural equation models including growth analysis, and Bayesian modeling/Bayesian Network building to CROM. He is also an experienced user, programmer, and trainer for computer and statistical programs such as SPSS, BayesiaLab, SAS, Matlab, EQS, Mplus, WINSTEPS, Conquest, BILOG, Parscale, WinBUGS,R, SQL/Hive, Spark, WinBugs, and KNIME. Dr. Yumoto’s programming skill allows him to conduct statistical analysis beyond the capability of pre-existing statistical software, which combines with his ability to integrate statistical and clinical considerations in the design, execution, and interpretation of research involving outcomes and metrics to greatly strengthen this Collaborative. Dr. Yumoto’s primary research interests are in design and development of appropriate assessment, bringing quantitative techniques from strictly psychometric contexts to practice, and techniques from machine learning and data science to assessment. He has also focused on improving the fairness of evaluations in longitudinal modeling applications. In his most recent evaluation project, he investigated the potential impact of unaccounted-for heterogeneity at the level of student growth on the estimation of teacher effects. He is also interested in applying psychometric methods to issues and problems in clinical psychology. He has been collaborating with a clinical psychologist on a longitudinal, multi-level model of post traumatic stress disorder since 2009, and with Dr. Tractenberg on longitudinal modeling – including establishing measurement invariance over time, incorporating biomarkers in multi-dimensional study endpoints, and building and validating Bayesian networks, relating to Alzheimer’s disease and brain aging – since 2006. Since CROM was established in 2008, Dr.s Yumoto and Tractenberg have coauthored six published articles with one more currently undergoing peer review, and they have collaborated on five posters at national and international meetings, and 11 talks.

Leave a Reply

Your email address will not be published. Required fields are marked *