Jody Schimmel Hyde, PhD, is a Research Scientist in the Survey Research Center (SRC) of the Institute for Social Research (ISR) at the University of Michigan (UM). She is a co-investigator and an associate director of the Health and Retirement Study. Dr. Schimmel Hyde’s research focuses on financial independence, employment, and public programs to support self-sufficiency and community living for people with disabilities and older adults. She is interested in survey measurement of disability, and the value of survey-administrative data linkages to understand participation in and outcomes for participants of income support programs. Prior to joining SRC, Dr. Schimmel Hyde was a Senior Fellow at Mathematica and the managing director of its Center for Studying Disability Policy. While at Mathematica, she was involved in research projects for federal agencies including the Social Security Administration, the Administration for Children and Families, the Administration for Community Living, and the Centers for Medicare & Medicaid Services.
Hoda Rahmani is a Postdoctoral Research Fellow at the Institute for Research on Innovation and Science (IRIS) at the University of Michigan. She earned her PhD and Master’s degrees in Industrial and Systems Engineering from Ohio University, with research focusing on applying machine learning and optimization techniques to complex systems.
This project aims to address the need for robust, rigorous and computationally efficient methods for optimizing Just-In-Time Adaptive Interventions (JITAIs) to prevent and treat substance use disorders (SUD). Although the proposed methods are motivated by micro randomized trials (MRTs) in SUD, they can be extended to observational studies and employed to develop effective JITAIs in other health domains as well. The methods developed in this project will help SUD scientists to more effectively leverage emerging technologies (e.g., mobile and wearable devices) to deliver support in a timely and ecological manner.
The New York City Housing and Vacancy Survey (NYCHVS) is a citywide, representative survey of New York City’s housing stock and population conducted about every three years. The NYCHVS collects data in neighborhoods, in all five boroughs, and focuses on representing all New Yorkers—regardless of who they are or where they live. All of this ensures that the NYCHVS represents the diversity of New York City’s residents and housing.
The NYCHVS selects 15,000 specific addresses throughout New York City to represent the more than 8 million residents who call New York City home. NYCHVS data are used to capture the diversity of New York City. Since the NYCHVS interviews collect the same information on all New Yorkers, everyone selected has an equal voice and opportunity to be heard.
The proposed study aims to implement a pilot project in the City of Detroit (n=1,200) with a novel, yet practical sampling approach designed to maximize coverage of microgeographies (Census block groups in our case) as well as inclusivity of groups who may be underrepresented in traditional sample surveys due to, for example, nonresponse and may hold different perceptions of community safety and policing (e.g., younger population) by blending address-based sampling (ABS) and respondent driven sampling (RDS). The sampling approach will rely on the principles of probability sampling, scientifically guaranteeing the representativeness. The various components in the hybrid sample will be blended through our extension of multiple frame estimation methods to fit our unique sample design as well as our major modification to the existing RDS estimators.
Cognitive Processing Therapy (CPT) and Prolonged Exposure (PE) are the most effective treatments for PTSD; however, up to 41% of patients do not improve significantly. It remains unclear how best to adapt treatment when patients show early non-response or struggle with between-session homework. To address this, our team developed the Hybrid Experimental Design (HED), a novel approach that integrates therapist-delivered and digital strategies. HED allows us to identify optimal treatment sequences for each patient. In our study (N=302), participants are randomized to CPT or PE. After 6 sessions, early responders or non-responders are further randomized to adapted treatment or maintenance strategies. Additionally, daily digital prompts are used to support homework completion. This research aims to develop more effective, efficient, and individualized PTSD treatment.
Randomized controlled trials (RCTs) are the gold standard for assessing interventions for preventing and treating cancer, but their external validity is only guaranteed if the trial participants are a random sample from the target population. Unfortunately, most cancer-related RCTs use convenience samples, not probability samples, and differences between the trial sample and the target population are likely to exist. If these differences are related to the effectiveness of the treatment being studied (effect modifiers), trial results will fail to generalize. While observable differences may be assessed and potentially adjusted for (e.g., underrepresentation of certain demographic groups), these differences have been shown to not completely explain the so-called efficacy-effectiveness gap. We posit that unmeasured differences between who chooses to participate in an RCT and who does not may be an important contributor to the failure of some trial results to generalize. In this project, we propose to develop a statistical framework for quantifying the potential impact of unmeasured differences between the trial sample and the target population on trial results. The resulting sensitivity analysis will bound the potential bias in the treatment effect estimate when generalizing from the trial sample to a target population. The methodology will be based on our prior work developing sensitivity analyses in the areas of survey nonresponse and selection bias which similarly consider the issue of differences between who is in a study sample and who is not. This work will have broad applicability beyond cancer trials, as generalizability is a universal concern of randomized trials across application areas.
Alzheimers Disease and Related Dementias (ADRD), set as a public health priority by the World Health Organization, is one of the leading causes of death in the U.S.; however, a dearth of ADRD research data that consider dynamics across racial/ethnic groups beyond minimum standard racial/ethnic categories is detrimental to understanding a comprehensive picture of the current state of ADRD and its future. Hybrid sampling (HybS) that combines address-based sampling and respondent driven sampling conducted through the push-to-Web method is an attractive and practical methodological option for yielding a probability sample for racial/ethnic minority data. This project aims 1) to establish an inclusive probability sample national survey panel of the general population of adults ages = 40 years old with an oversample of Asian Americans focusing on the granular subgroups of Chinese, Asian Indian, Filipino, Korean and Vietnamese, and of Latinx stratified by Afro-heritage using the push-to-Web HybS and 2) to study ADRD risks across racial/ethnic groups and the role of social networks on ADRD risks.
This project aims to further advance capabilities in the social sciences (broadly defined) to collect data on the daily lives of families and individuals. These data will be more accurate, more granular, and more complete than has been possible to obtain in traditional survey-based research until now. The context for this is the Understanding America Study (UAS), the probability-based Internet panel we have been building at USC since 2014. The infrastructure includes the combination of many data types (including survey data, information collected from wearables, contextual data, administrative linkages, ecological momentary assessments, self-recorded narratives, and electronic records of financial transactions), as well as an open communication with the wider research community both in making data available as quickly as possible and in soliciting input about content and methods.
Family caregivers are essential to the nations well-being and economy yet little information exists regarding the daily lives of Alzheimer’s Disease and Related Dementias (ADRD) and non-ADRD caregivers across the diverse social partners who provide care.
The present study develops and administers new survey instruments in the Understanding America Study (UAS) (https://uasdata.usc.edu/index.php) to identify caregivers and implements new daily assessments and wearable devices to capture caregivers daily experiences and their links with daily emotional and physiological well-being among ADRD and non-ADRD family and non-family caregivers ranging in age from young adulthood to old age.
Understanding daily experiences and reactivity and the factors that predict greater resilience and vulnerability to stress in a national sample of caregivers will provide unprecedented information regarding potentially modifiable risk and protective factors for improving caregiver health and well-being.