Examples of four engagements across defense, government, and industry — each one built to answer a question existing methods couldn't reach. Together they trace a career spent turning applied psychology into decisions leaders could act on.
The Navy's ability to compete for talent depends on how well it develops, evaluates, and selects its people. Michael led teams for evidence-based research and development across three coordinated initiatives building an integrated, modernized system focused on Sailor competencies necessary for mission success rather than that mere administrative compliance.
MyNavy Coaching fostered a culture of peer-to-peer developmental conversations where Sailors identify and own their own developmental success.
Navy Performance Management Administration (NPMA) -- a digital performance management system steeped in evidence-based research and best practices from both Government and industry -- is in prototype form and poised to replace paper-based evaluations with a foundation for fairer, clearer performance data.
The Warrior Development Program (WDP) modernizes Command Qualification processes to select our best and brightest leaders through a data-driven understanding of a candidate’s leadership competencies including leadership attributes, character, potential, and community-specific values.
Three initiatives, one integrated strategy: a Navy that develops talent more deliberately, selects leaders more accurately, and retains its best people.
The insight beneath the model: turnover isn't just a work problem -- work and nonwork life needs dynamically spill into each other, and the overlap predicts who leaves. That finding became SIMPACT, a decision-support tool that identifies which life needs are most strongly tied to turnover intent and lets organizations run predictive "what if" simulations before spending a dollar on employee retention programs for targeted employee retention investment.
A validated, whole-person turnover model — extended from military quality-of-life research into a tool public- and private-sector employers could use directly.
A multi-faceted National Cancer Institute Study examined radiation exposure for those that lived in villages near or around the Semipalatinsk Nuclear Test Site in the 1950s. These studies are critical in understanding the effects of internal and external radiation exposure on people and how those exposures affected health outcomes. Existing data overlooked culturally critical dietary consumption practices at the time and the data needed didn't exist and could not be collected again with aging, rural village residents. The research team returned to affected villages using focus group interview and key informant interview methodologies Dr. Schwerin developed to attempt to reconstruct quantitative data using a qualitative focus group interview methodology. Focus group facilitators were trained and data collection tools were developed in two languages - Kazakh and Russian - for new estimates of dairy consumption and daily life in 1950s Kazakhstan. These novel methods surfaced a critical, previously underestimated detail: 1950s ethnic Russian and Kazakh families used a wider variety of animal milk and milk products than previously believed including horse, sheep, goat milk, and camel milk. This distinction mattered in the radiation exposure models because radiation was held in milk fat and the amount of milk fat varied greatly by the animal it was sourced from.
Corrected assumptions about diet, housing, and time spent outdoors materially improved the radiation dose estimates behind a major public-health study. The method itself became a template other historical-exposure studies could reuse wherever written records ran out.
Terrorist and extremist groups don't produce clean data for modeling but publicly available data on extremist groups and violent events hold fragments of evidence about their history of violence, ideology, capabilities, and attack planning. Working through the Institute for Homeland Security Solutions for DHS's Science and Technology Directorate, the research elicited theory-based models from subject matter experts, applied social-science-based quantitative models trained and validated on historical data from violent and nonviolent groups alike to a genuinely sparse evidence problem.
The resulting tool runs a three-step process to examine alternate, competing hypotheses: frame the question and characterize the group, apply quantitative models grounded in recognized research, then explain the results by tracing them back to specific pieces of evidence including what additional information would most reduce uncertainty.
An analytical tool that gives analysts alternative, evidence-backed hypotheses about a group's intentions — and tells them exactly where to look next.