making decisions for complex, dynamic problems with imperfect knowledge: the application of control...

49
Making decisions for complex, dynamic problems with imperfect knowledge The application of control systems engineering to a behavioral intervention @ehekler Eric Hekler, PhD Arizona State University August 18, 2016 Flickr -Pat Castald 1

Upload: designing-health-lab-arizona-state-university

Post on 11-Apr-2017

290 views

Category:

Science


0 download

TRANSCRIPT

Behavioral Theory A Primer about concepts and behavior-change techniques

Making decisions for complex, dynamic problems with imperfect knowledge The application of control systems engineeringto a behavioral intervention@eheklerEric Hekler, PhDArizona State UniversityAugust 18, 2016

Flickr -Pat Castaldo1

The talk will briefly set up the current context for mHealth / UbiComp / digital health research efforts as seen from various disciplinary lenses. Following this, the precision medicine initiative will be discussed followed by a discussion on one subclass of prevention interventions, labeled precision behavior change, which could fit well within the precision medicine initiative. Following the definition of precision behavior change, transdisciplinary research questions, with a particular focus on attempting to articulate intellectual merit and contributions for each discipline when exploring the research questions, will be discussed. The talk will conclude with plausible next steps to spur conversation among the webinar participants and later viewers on ways to refine this transdisciplinary research agenda to see if it is viable and, if so, how best to more actively enable it as an organizing moon shot agenda for the mHealth research community.1

A group effort

2@ehekler

2

OutlineEpistemological targetControl Systems EngineeringOur workEncapsulate previous knowledgeDefine dynamic decisions of an interventionDevise a system ID experimentExamine individual differences (ARX)Examine mechanistic model (semi-physical modeling)Devise model-predictive controller

@ehekler3

The talk will briefly set up the current context for mHealth / UbiComp / digital health research efforts as seen from various disciplinary lenses. Following this, the precision medicine initiative will be discussed followed by a discussion on one subclass of prevention interventions, labeled precision behavior change, which could fit well within the precision medicine initiative. Following the definition of precision behavior change, transdisciplinary research questions, with a particular focus on attempting to articulate intellectual merit and contributions for each discipline when exploring the research questions, will be discussed. The talk will conclude with plausible next steps to spur conversation among the webinar participants and later viewers on ways to refine this transdisciplinary research agenda to see if it is viable and, if so, how best to more actively enable it as an organizing moon shot agenda for the mHealth research community.

3

From generally useful to useful for me

Epistemological target4@ehekler

4

Embracing (plausibly) meaningful variability@eheklerIn General~50%Personalization/Matchmaking~35%Idiosyncratic/Subjective~15%Hekler, et al. 2016, Agile Science, Translational Behavioral Medicine5

Professionals still focus on on average science (even, it appears, with many precision medicine efforts)Professionals need to move towards studying the utility of personalization algorithms

5

Specific Solutionsfor Specific Problems

Design & Engineering

On Average ScienceOn Average Evidencefor General ProblemsKey

Traditional pathway

Emerging pathway

Product

ProcessProfessional-led6@ehekler

Decision Policies we are talking about what this is supposed to do

Citizens= Patients, Providers, and anyone else driven to solve a problem that the individualhas first-hand experience with.

6

Specific Solutionsfor Specific Problems

Design & Engineering

On Average ScienceOn Average Evidencefor General ProblemsKey

Traditional pathway

Emerging pathway

Product

ProcessPrecise Evidencefor Specific Problems

Personalization AlgorithmScienceProfessional-led

Process IndividualizationScience7@ehekler

Decision Policies we are talking about what this is supposed to do

Citizens= Patients, Providers, and anyone else driven to solve a problem that the individualhas first-hand experience with.

7

Specific Solutionsfor Specific Problems

Design & Engineering

On Average ScienceOn Average Evidencefor General ProblemsKey

Traditional pathway

Emerging pathway

Product

ProcessPrecise Evidencefor Specific Problems

Personalization AlgorithmScienceProfessional-led

Process IndividualizationScienceCitizen/Patient-led8@ehekler

Decision Policies we are talking about what this is supposed to do

Citizens= Patients, Providers, and anyone else driven to solve a problem that the individualhas first-hand experience with.

8

Specific Solutionsfor Specific Problems

Design & Engineering

On Average ScienceOn Average Evidencefor General ProblemsKey

Traditional pathway

Emerging pathway

Product

ProcessPrecise Evidencefor Specific Problems

Personalization AlgorithmScienceProfessional-led

Process IndividualizationScience

9@ehekler

Decision Policies we are talking about what this is supposed to do

Citizens= Patients, Providers, and anyone else driven to solve a problem that the individualhas first-hand experience with.

9

Making decisions in complex, dynamic systems with imperfect knowledge

Control Systems Engineering10@ehekler

10

Control Systems EngineeringNSF IIS-1449751: EAGER: Defining a Dynamical Behavioral Model to Support a Just in Time Adaptive Intervention, PIs, Hekler & Rivera@ehekler11

11

Describe & predict: System identificationNSF IIS-1449751: Defining a Dynamical Behavioral Model to Support a Just in Time Adaptive Intervention, PIs, Hekler & Rivera@ehekler12

Myc olleauge, Daniel Rivera, and I have been extending this further using methods fromcontrol systems engineering to develop experimental designs that take more advantage of a priori knowledge than the micro-randomization study. In the discussion section, Id be happy to get into details on these experimental designsbut for the focus of this, the main point is to realize that this is a huge shift in the behavioral science community away from ideas like RCTs nad instead towards methods that embrace and map out idiosyncracy.12

Martin, Rivera, & Hekler Am. Control Conference (2015)

Control: Model-predictive control@ehekler13

Based on this, we need to move more into an open discussion in which we explore lots and lots of different ideas if we really want to understand which ones are best.Sadly, science, particularly behavioral science doesnt really have the sort of maker culture that would allow us. As such, a key emphasis. 13

Continuous improvement: Adaptive control@ehekler

Flickr - Dave Gray14

Systematically managing and mitigating imperfect knowledge to support dynamic evidence-based decisions

Our work15@ehekler

15

Encapsulate previous knowledge (theory)@ehekler16

Dynamical model of Social Cognitive TheoryRiley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014

@ehekler17

One inventoryRiley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler

18

Differential equations (first order shown)Riley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler

19

Simulation: Low vs. high self-efficacyRiley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler

Low Self-EfficacyHigh Self-Efficacy20

Simulation: HabituationRiley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler

21

Secondary data analysis: ValidationRiley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler

22

Define dynamic decision(s) of intervention@ehekler23

Daily ambitious but doable step goalsHekler (PI), Rivera (Co-PI), NSF IIS-1449751

@ehekler24

Intervention decisions

Martin, Rivera, & Hekler, 2015; 2016, American Control Conference25@ehekler

Simulation

Martin, Rivera, & Hekler, 2016, American Control Conference26@ehekler

Devise a system identification experiment@eheklerNeed: replicable (for estimation & validation), random/pseudo-random, excitation over time27

Experimental design: System identification

28@ehekler

Coming to how the daily goal signal was designed:

Then developed an experimental design In control systems world, this methodology is called system identification. It is to test this hypothesis, Estimate and validate the dynamical model.Focus is on idiographic modeling, individual model per participant.

System ID experiments are specifically designed to estimate and validate the dynamical model, and the focus is on idiographic models, meaning individual models per participant or user. Every day, a step goal (external cues), and points is assigned to the participant (outcome expectancy for reinforcement). Step goals range from doable (baseline median), to ambitious (up to 2.5x baseline).Each individual has her own unique randomization signalThis strategy also uses cycles of the intervention, for us, we used 16 day cycles. So the same randomization signal repeats every 16 days.The randomization signal is determined using multisine wave design strategies, which maximizes the signal to noise, delievered orthogonally in frequency, and is useful for progressively testing model fit, thus making it valuable for understanding how dynamics change over time for an individual.

Multisine signal design utilizes periodic signals defined in the frequency domain to implement an open-loop experiment (see C.2.1). A useful analogy is an audio equalizer whereby different frequencies like bass or treble can be emphasized; frequencies occurring as cycles across time can be used to design an experimental signal (e.g., daily goal variations).

One thing to note here, this was not a perpetually adaptive or personalized intervention, it was mainly designed to understand the dynamics and build individualized computational models.

Pseudo-randomly assigns daily goals and points to every participant 28

Multisine pseudo-random signals

Martin, Rivera, & Hekler, 2015, American Control Conference29@ehekler

Pilot study: Just Walk

Fitbit Zip

30@ehekler

We ran our pilot study from June to December 2015, and the quick summary so that everything makes sense is that the study was 14 weeks long, participants received a Fitbit, an Android app, and a daily step goal and we measured many contextual variables that were informed from the SCTon a daily/weekly/monthly level depending on the variables.

Just Walk is the system that we developed for running the experiment.

It includes a front-end Android appNotifications of daily step goal and corresponding pointsNotifications when user achieves goalsIntegration with Fitbit which is used to measure the steps

30

Black box modeling to develop descriptive models & examine individual differences@ehekler31

Participants22 inactive, overweight Android usersBMI 33.7 6.747 6.2 years 87% womenLiving anywhere in the USAverage Baseline Median Steps: 4972 steps/day (SE = 482)

32@ehekler

Daily Morning and evening self-report We also collected weather, and location data but have not analysed that yet.

We recruited 22 inactive, overweight Android users (one lost her Fitibit during last three weeks but was willing to continue but this compromised system ID analyses for her as our power calculations required a minimum of 5 cycles; final sample N=21; 90% women; M = 47.0 6.2 years, BMI 33.7 6.7). Baseline median steps averaged 4,972 steps/day (SE = 482), and median steps in the last cycle were 6,827 steps/day (SE = 647). By design, there was an average 45% (SD = 36%) increase in steps/day from baseline to the last cycle, and participants met 69% (SD = 24%) of goals. Results from a nonlinear mixed effects model indicated a significant, on average, increase in steps from baseline to the first intervention cycle of 1,500 steps (t=-5.52, p