Near all of my students are from the Department of Statistics; a small minority of students who work with me come from other departments at UMich, including the Departments of Biostatistics, Psychiatry, Learning Health Sciences. All students who are interested in learning more about my statistical or methodological research are (1) encouraged to visit the d3c website, (2) strongly encouraged to reach out to any of my current and prior students or trainees, and are (2) welcome to attend any of our public activities at d3c, including Think Tank sessions and our weekly Friday center/lab meetings. Consult the d3c Events Calendar for more information
PhD Students. Currently, I am not accepting new PhD Students.
Masters Students. I no longer accept Masters students from the Dept of Statistics.
Undergraduate Students. Currently, I am not accepting new Undergraduate students.
Scroll down to Scroll down to here if you are interested in working with me and if I am currently accepting students like you.
Welcome
Welcome to my homepage. I'm an Associate Professor in the Survey Research Center of the Institute for Social Research (ISR) and in the Department of Statistics at the University of Michigan. I am also Co-Director of the Data Science for Dynamic Intervention Decision-making Center (previously d3lab, also known as d3c or d-cubed) within the Quantitative Methodology Program at the ISR. Since 2002, I had been affiliated with the Methodology Center at Penn State University; the Methodology Center closed its doors around 2019. I have a Ph.D. in Statistics from the Department of Statistics at the University of Michigan (class of 2007). Prior to coming back to Michigan as Faculty, from 2007 and 2009, I was a Research Investigator in the Durham VA Center for Health Services Research and Development in Primary Care (HSR&D), and an Assistant Professor in the Department of Biostatistics and Bioinformatics at Duke University.
I am a statistician, methodologist, and intervention scientist. I spend all of my time at work researching and developing tools that can be used to (learn how best to) improve health, education and well-being. Broadly speaking, the tools I develop or co-develop fall into one of two categories: new approaches to data collection (primarily, different types of randomized trials) and new approaches to data analysis. I develop these tools primarily for use by other researchers (e.g., psychiatrists, psychologists, education/behavioral scientists or other data scientists) who are developing new interventions for improving health, education and well-being. Often, I work closely with these other scientists to directly apply the methods I develop. In 2012, I co-founded d3lab, now d3c, with my long-time colleague and friend Dr. Inbal Nahum-Shani. d3c is a growing community of senior and junior scientists, postdoctoral fellows, and graduate and undergraduate students with whom I collaborate. I also enjoy mentoring the next generation of statisticians and data scientists: my mentees include undergraduate and graduate students in the Department of Statistics, as well as postdoctoral and early career investigators across a wide variety of health and education research areas.
My specific methodological interests are varied, but they all involve work at the intersection of Statistics and Intervention Science. I am particularly interested in developing statistical methods that can be used to form
adaptive interventions, sometimes known as dynamic
treatment regimes. An adaptive intervention is a sequence of individually tailored decisions rules
that specify whether, how, or when--and importantly, based on which measures--to alter the intensity, type, or delivery of treatment at critical decision points during intervention.
Adaptive interventions are particularly well-suited for the management of chronic diseases, but can be used in any clinical or educational setting in which sequential
medical decision making is essential for the welfare of the individual. They hold the promise of enhancing clinical practice by flexibly tailoring treatments or interventions to individuals when they need it most,
and in the most appropriate dose, thereby improving the efficacy and effectiveness of treatment. In health settings, adaptive interventions represent one important tool in the practice of "precision medicine". However, adaptive interventions can also be used to adapt interventions at the organizational level, for example, to encourage clinics or schools to adopt an evidence-based intervention.
I devote a great portion of my time to addressing methodological issues in the design of sequential multiple assignment randomized trials (SMARTs), and other randomized trial designs, that can be used to optimize or evaluate adaptive interventions.
My two areas of application are health care and education. The methods I develop and co-develop can be applied across a wide variety of areas. I am particularly interested in their application in the substantive areas of mental health (e.g., autism, depression, anxiety) and substance abuse, especially as related to children and adolescents.
Key Words: dynamic treatment regimes, adaptive treatment strategies, sequential multiple assignment randomized trials, adaptive implementation interventions, causal inference, propensity score methods, marginal and structural nested mean models, methods for longitudinal data analysis, health services research, mental health, substance abuse, obesity
Below is a small selection of my published research that does not get updated as of 2020 (a * means this is a student I mentored). For a more complete and updated list of my publications, please access my CV and/or see my Google Scholar page.
*Luers, B., Qian. M., Nahum-Shani. I., Kasari. C. and Almirall D. (under review). Longitudinal Mixed Models for Comparing Embedded Dynamic Treatment Regimens in Sequentially Randomized Trials. [arXiv version].
Quanbeck, A., Almirall, D., Jacobson, N., Brown, R. T., Landeck, J. K., Madden, L., Cohen, A., Deyo, B.F., Robinson, J., Johnson, R. and Schumacher, N. (2020). The Balanced Opioid Initiative: protocol for a clustered, sequential, multiple-assignment randomized trial to construct an adaptive implementation strategy to improve guideline-concordant opioid prescribing in primary care.Implementation Science, 15, 1-13 [Article 20] [Technical supplement with pre-specified analysis plan].
Nahum-Shani. I. and Almirall D. (2020). An Introduction to Adaptive Interventions and SMART Designs in Education.NCSER 2020-001, US Department of Education. Washington, DC: National Center for Special Education Research. [Retrieved December 1, 2019 from https://ies.ed.gov/ncser/pubs/]
This manuscript was commissioned by the US Department of Education, and was peer reviewed by an independent panel of the Standards and Review Office of the Institute of Education Sciences. Though not in a scholarly journal, it is considered a major publication.
*Seewald, N., Kidwell, K., Nahum-Shani, I., Wu, T., McKay, J., and Almirall D. (2019). Sample size considerations for comparing dynamic treatment regimens in a sequential multiple-assignment randomized trial with a continuous longitudinal outcome.Statistical Methods in Medical Research [arXiv version].
Almirall D., Nahum-Shani, I., Wang, L., Kasari, C. (2018). Experimental Designs for Research on Adaptive Interventions: Singly- and Sequentially-Randomized Trials.
*Hall, K., Nahum-Shani, I., August, G., Patrick, M., Murphy, S.A., Almirall D. (2018). Adaptive Prevention Designs in Substance Use Prevention.
Boruvka, A., Almirall D., Murphy, S.A. (2017). Assessing Time-Varying Causal Effect Moderation in Mobile Health: Modeling and Estimation Considerations for Intensive Longitudinal Intervention Data. [arXiv version] Journal of the American Statistical Association.
*NeCamp, T., Kilbourne, A. Almirall D. (2017). Cluster-level adaptive interventions and sequential, multiple assignment, randomized trials: Estimation and sample size considerations.Statistical Methods in Medical Research
Almirall D., Chronis-Tuscano, A., (2016). Adaptive interventions in Child and Adolescent Mental Health.Journal of Clinical Child and Adolescent Psychology.
*Hwanwoo, K., Ionides, E., Almirall D. (2016). A Sample Size Calculator for SMART Pilot Studies.Society for Industrial and Applied Mathematics (SIAM): Undergraduate Research Online (SIURO) Journal.
*Lu, X., Nahum-Shani, I., Kasari, C., Lynch, K.G., Oslin, D.W., Pelham, W.E., Fabiano, G., Almirall D. (2015). Comparing dynamic treatment regimes using repeated-measures outcomes: modeling considerations in SMART studiesStatistics in Medicine. [Abstract] [Technical Report of an older version of the manuscript is available from The Methodology Center at Penn State University]
Chronis-Tuscano, A., Wang, C.H., Strickland, J., Almirall D., Stein, M.A. (2016). Moving Toward Personalized Treatment of Mothers with ADHD and Their At-Risk Children: A SMART Pilot.Journal of Clinical Child and Adolescent Psychology.
Almirall D., DiStefano, C., Chang, Y., Shire, S., Lu, X., Nahum-Shani, I., Kasari, C. (2016). Adaptive interventions and longitudinal outcomes in minimally verbal children with ASD: Role of speech-generating devices.Journal of Clinical Child and Adolescent Psychology.
This manuscript was recognized as a "top 20 article" in Autism in 2016 by the Interagency Autism Coordinating Committee; see here.
Gunlicks-Stoessel, M., Mufson, L., Westervelt, A., Almirall D., Murphy S.A. (2014). A Pilot SMART for Developing an Adaptive Treatment Strategy for Adolescent Depression.Journal of Clinical Child and Adolescent Psychology.
Kilbourne, A.M., Almirall, D., Eisenberg, D., Waxmonsky, J., Goodrich, D.E., Fortney, J.C., Kirchner, J.E., Solberg, L.I., Main, D., Bauer, M.S., Kyle, J., Murphy, S.A., Nord, K.M., Thomas, M.R. (2014). Protocol: Adaptive Implementation of Effective Programs Trial (ADEPT): cluster randomized SMART trial comparing a standard versus enhanced implementation strategy to improve outcomes of a mood disorders program.Implementation Science, 9,132. DOI: 10.1186/s13012-014-0132-x.
Almirall D., Nahum-Shani, I., Sherwood, N.E., Murphy S.A. (2014). Introduction to SMART Designs for the Development of Adaptive Interventions: With Application to Weight Loss Research.Translational Behavioral Medicine, 4:260-274. DOI: 10.1007/s13142-014-0265-0 [Technical Report of an older, longer, and not-as-polished version of this manuscript is available from The Methodology Center at Penn State University]
Kasari, C., Kaiser, A., Goods, K., Nietfeld, J., Mathy, P., Landa, R., Murphy, S.A., Almirall D (2014). Communication Interventions for Minimally Verbal Children with Autism: Sequential Multiple Assignment Randomized Trial.Journal of the American Academy of Child and Adolescent Psychiatry. DOI:10.1016/j.jaac.2014.01.019
This manuscript was discussed by Dr. Helen Tager-Flushberg of Boston University; see here.
Almirall D., McCaffrey, D.F., Griffin B.A., Ramchand R., Yuen R., Murphy S.A. (2013). Time-varying effect moderation using the structural nested mean model: estimation using inverse-weighted regression-with-residuals.Statistics in Medicine. [Technical Report of an older version of the manuscript is available from The Methodology Center at Penn State University]
McCaffrey, D.F., Griffin, B.A., Almirall D., Slaughter, M.E., Ramchand, R., Burgette, L.F. (2013). A Tutorial on Propensity Score Estimation for Multiple Treatments using Generalized Boosted Models.Statistics in Medicine. DOI: 10.1002/sim.5753
Almirall D., Compton S.N., Rynn M.A., Walkup J.T., Murphy S.A. (2012). SMARTer Discontinuation Trial Designs for Developing an Adaptive Treatment Strategy.Journal of Child and Adolescent Psychopharmacology 22(5):364-74. DOI: 10.1089/cap.2011.0073. PMID: 23083023. PMCID: 3482379. [Technical Report of an older version of the manuscript is available from The Methodology Center at Penn State University]
Nahum-Shani I., Qian M., Almirall D., Pelham W.E., Gnagy B., Fabiano G., Waxmonsky J., Yu J., Murphy S.A. (2012). Experimental Design and Primary Data Analysis Methods for Comparing Adaptive Interventions.Psychological Methods. [Technical Report of an older version of the manuscript is available from The Methodology Center at Penn State University]
Nahum-Shani I., Qian M., Almirall D., Pelham W.E., Gnagy B., Fabiano G., Waxmonsky J., Yu J., Murphy S.A. (2012). Q-Learning: A Data Analysis Method for Comparing Adaptive Interventions.Psychological Methods. [Technical Report of an older version of the manuscript is available from The Methodology Center at Penn State University]
Almirall D., Lizotte, D., Murphy S.A. (2012). SMART Design Issues and the Consideration of Opposing Outcomes, a Discussion of Evaluation of Viable Dynamic Treatment Regimes in a Sequentially Randomized Trial of Advanced Prostate Cancer by Wang, Rotnitzky, Lin, Millikan, and Thall. Journal of the American Statistical Association (Case Studies and Applications) 107(498):509-12.
Almirall D., Compton S.N., Gunlicks-Stoessel M., Duan N., Murphy S.A. (2012). Designing a Pilot Sequential Multiple Assignment Randomized Trial for Developing an Adaptive Treatment Strategy.Statistics in Medicine 31(17):1887-902. DOI: 10.1002/sim.4512. PMID: 22438190. PMCID: 3399974. [Technical Report of an older version of the manuscript is available from The Methodology Center at Penn State University]
Almirall D., McCaffrey, D.F., Ramchand R., Murphy S.A. (2013). Subgroups Analysis when Treatment and Moderators are Time-varying.Prevention Science, 14(2):169-78.
Almirall D., Ten Have T., Murphy S.A. (2010). Structural Nested Mean Models for Assessing Time-Varying Effect Moderation.Biometrics, 66(1):131-139.
Presentations
Slides of my presentations can be found here; if you click on "Last modified", the files will sort by date (eventually, I will organize the files for easier viewing and download).
Old workshop slides (no longer being updated) on the topic of adaptive interventions and sequential multiple assignment randomized trials (SMART) can be found here.
Newer and more up to date slide decks and related resources can be found under the Resources tab of our d3c website.
First, please scroll up to the top of this webpage and make sure I am currently accepting new students. I only have the capacity to work with a limited number of students at a time. Second, if I am accepting new students like you, please read the following carefully before sending me an email. Much of the text below is copied from the website of Finale Doshi-Velez, a computer scientist at Harvard doing some interesting research at the intersection of machine learning and healthcare!
For undergraduate students: Experience with (i) some kind of programming (e.g., R, Matlab, Fortran, C++ or python) and (ii) regression analysis is a must. I ask for at least a one year commitment; though note that I prefer at least a 2.5 year commitment as I find this has led to the most productive and impactful projects. The project must lead to an undergraduate thesis or honors thesis; though note that I prefer to work with students who are interested in publishing their work in a scholarly journal (this includes undergaduate research journals, for example, like this). I have a strong preference for 2 undergraduate students, or 1 undergraduate together with a graduate student, to work on a project together. I only advise students on projects that are related to the core directions of my lab (i.e., dynamic intervention decision-making, see above).
For PhD students: Experience (or graduate coursework) in (i) probability/statistics (at least an introduction to mathematical statistics, e.g., using the Delta method to derive the limiting distribution of an estimator), (ii) linear algebra, and (iii) programming (e.g., R, Matlab, Fortran, C++ or Python) is a must. Experience or strong interest in (iv) causal inference will lead to greater productivity (but note that this is not a deal-breaker). I ask for at least a 3 year commitment; though note that I prefer at least a 4 year commitment as I find this has led to the most productive and impactful projects. The project must lead to publication in a Statistics or Biostatistics journal. I only advise students on projects that are related to the core directions of my lab (i.e., dynamic intervention decision-making in health and education settings, see above).
Expectations of all students:Expectations in terms of technical skills: I expect all of my students to have (or be interested in having) excellent communication skills (written and spoken). I expect all of my students to write computer software independently (i.e., I'm not going to be teaching you how to code -- in fact, you will be teaching me:). Expectations in terms of non-technical skills: More than any set of specific technical skills, however, I'm looking for students with (i) an interest in sharpening their curiosity (excitement about the question, more than about the answer), (ii) an interest in risk-taking and learning from failures, (iii) an interest in persevering when tackling tough, meaningful problems and (iv) the ability to receive and process constructive criticism (tough love) without falling apart. If you're reaching out to me, feel fee to include a personal story of working through something really exciting and really hard, which may or may not have worked out. I expect my students to be organized and timely with (at least) weekly updates via in-person meeting, email, or phone. I expect my students to communicate challenges so we can work through them together. I expect my students to lift up others around them. If you do not identify as a hard worker or you do not want to learn how to work independently, then please do not contact me.
Steps to getting in touch with me via email. If you made it this far and you think I would make a good mentor, do the following: Please read at least 1 of my manuscripts published in the last 5 years (this can be one where I am first, second, or last author -- typically I am last author when I publish with my students). Please get to know the d3c website. Please let me know specifically why you are interested in working in our d3c or with me. Please show me that you've read the text above by structuring your email subject line as "Lento pero aplastante: [ FILL IN YOUR SUBJECT LINE ]" and be sure to CC Phil Stranyak (stranyak@umich.edu) in your email to me.
L to R: Inbal "Billie" Nahum-Shani, Daniel Almirall, Susan A. Murphy in the Fall 2015
Statistical Reinforcement Learning Lab of the Quantitative Methodology Program, Survey Research Center, Institute for Social Research, Fall 2015.
At the UCLA Kasari Autism Research Lab, February 2015.
L to R: Charlotte DiStefano, Wendy Shih, Ya-Chih "Jilly" Chang, Ansel Almirall (son), me, Connie Kasari, Stephanie Shire.
L to R: Olivia Hackworth, Brook Luers, Tim NeCamp in Spring 2018.
Some of the members of our lab in Spring 2018.
First Published: 01/05/2010; Last Revised: 6/16/2020
