skip to Main Content

Data Science for Dynamic Intervention Decision-making Lab

Getting SMART About Adaptive Interventions in Education

A four-day workshop on adaptive interventions and sequential multiple-assignment randomized trials (SMART) in education.

 

Date of Training
March 11-14, 2019

 

Application Period
October 1, 2018 to 11:59PM (EST) December 16, 2018. Applicants will be notified about decisions by January 31, 2019.

 

Who Should Apply?

Anyone with a doctoral degree who is interested in learning more about adaptive interventions in education and in the use of sequentially-randomized trials to build high-quality adaptive interventions is encouraged to apply.

 

Prerequisites

Each scholar should: (a) have a graduate doctoral degree; (b) be working in early intervention, education, or special education; and (c) be a citizen or non-citizen national of the United States, or must have been lawfully admitted to the United States for permanent residence (i.e., valid I-551, or other legal verification of such status).

 

Fee or Costs

There is no registration fee. However, scholars who are accepted must arrange for their own travel expenses (e.g., transportation, lodging in Ann Arbor). A limited number of travel scholarships will be awarded to scholars who are unable to fund their own travel; scholars can apply for scholarships after they are notified of their selection.

 

Request additional information

The application portal is now open.

Objective

The goal of this four-day training is to promote ongoing professional development among education scientists interested in conducting research on adaptive interventions. This includes training in the design, conduct and analysis of sequential multiple assignment randomized trials (SMART).

 

Format

Time will be spent in lecture, discussion, Q&A, brainstorming sessions, small group work (practicum) and software demonstration. Lectures are provided by methodological scientists who are experts in research on adaptive interventions, and by intervention scientists (guest experts) who have conducted studies of adaptive interventions in education.

 

Enrollment

Enrollment is limited to approximately 30 scholars.

 

Topics Covered

  • Adaptive Interventions
  • Sequential Multiple Assignment Randomized Trials
  • SMART Case Studies
  • Analysis of Data Arising from a SMART
  • Preparing for a SMART
  • Pilot SMART
  • Other Experimental Designs in Research on Adaptive Interventions

Full agenda to come.

The workshop will be led by Inbal Nahum-Shani and Daniel Almirall, Co-Directors of the d3lab: The Data Science for Dynamic Intervention Decision-making Lab at the University of Michigan. In addition, the workshop includes guest lectures by experts in research on adaptive interventions in education: Meredith Gunlicks-Stoessel (University of Minnesota), Connie Kasari (UCLA), William Pelham (Florida International University), and Gregory Roberts (University of Texas, Austin)

What is an adaptive intervention?

Adaptive interventions use a sequence of decision rules that guide whether, how, or when—and, importantly, based on which measures—to make critical decisions about interventions in education settings. This includes whether, how or when to alter the dosage (duration, frequency, or amount), type, or delivery of interventions to students (or organizations). These interventions seek to address the individual and changing needs of students (or organizations) as they progress through an intervention.

 

What is a Sequential Multiple Assignment Randomized Trial (SMART)?

A SMART is a type of multi-stage, experimental design that was developed explicitly for constructing effective adaptive interventions. In a SMART, some or all participants are randomized multiple times over the course of the study. The multiple, sequential randomizations in a SMART enable researchers to efficiently address multiple scientific questions concerning the selection and individualization of intervention options at various decision points of an AI.

 

Why is this workshop needed?

Despite the critical role adaptive interventions already play (and will continue to play) in various domains of education, experimental research aiming to systematically optimize adaptive interventions in education is still in its infancy. SMARTs are experimental designs that enable scientists to address multiple scientific questions for optimizing a high-quality AI, but because SMARTs are relatively new, most educational researchers have not been exposed to them as part of their formal training. While research on AIs and SMART methods has grown significantly in the past few years, there is currently no comprehensive training in AIs and SMARTs in education. This workshop attempts to fill this gap.

This training institute is supported by the Institute of Education Sciences, U.S. Department of Education, through Grant Number R324B180003 to the Regents of the University of Michigan.

Back To Top
Search