Broadly, my research focuses on applying new methods to improve the design and evaluation of nutrition behavior change interventions.

Dissertation

Big question

Can studying differences in implementation and context during a school-based nutrition intervention RCT provide actionable direction for program improvement?

Methods

1. Hierarchical Linear modeling
2. Mixed Methods
3. Qualitative Methods

Findings

• universally challenging food environment
• individual students responded differently to the same intervention
• students’ varying degrees of behavior change were linked to different psychological and social characteristics

Upshot

I identified four unique profiles, or “phenotypes,” of behavior change and related psychosocial mediators among these students. This research was awarded a student research award at the Society for Nutrition Education and Behavior annual conference and led to my pursuit of a precision medicine equivalent for behavioral nutrition interventions.

Current Research: Psychosocial Phenotyping

Big Question

How can computational methods help us identify psychosocial phenotypes?

Methods

In a project funded by a New York Academy of Sciences Sackler Institute for Nutrition Early Career Award, I aim to

1. identify the psychosocial phenotypes that exist among residents of Washington Heights, a predominantly Latino community in NYC with high levels of diabetes and health disparity
2. develop a mobile application that identifies individuals’ psychosocial phenotypes

Current Research: Mechanisms of Behavior Change

Big Question

What are the mechanisms by which we can use digital technology to intervene on mediators of nutrition behavior?

Methods

• experiments on a social computing platform, LabInTheWild, that engages online volunteers in behavioral research
• a nutrition knowledge test that asked participants to compare the macronutrient content of real meal photos

Results

On average, participants experienced measurable changes in nutrition knowledge when either an expert or another test-taker explained the correct answer

Read press from Harvard School of Engineering and Applied Sciences here

Upshot

These results suggest the potential for social computing as a setting for nutrition education. A second series of studies examining the influence of peer and expert explanations on attitudes about the healthfulness of controversial foods will launch Spring 2017 and aims to elucidate the spread of misinformation about nutrition and identify approaches for nutrition experts to debunk such myths. Findings from this research will help me select more effective intervention strategies when developing tailored nutrition interventions.

Research Trajectory

Hypothesis

Identifying psychosocial phenotypes with data science methods and linking them to specific mediators and defined mechanisms for behavior change will lead to more efficient and more personalized tailoring of theory-based nutrition interventions.

Upshot

In future research, I plan to use psychosocial phenotypes to select representative community members to participate in the user-centered design of a tailored nutrition intervention and assess the generalizability of these methods in other communities.