A History of STATCOM at the University of Michigan
The STATCOM chapter at the University of Michigan was founded in December 2006, after a visit from Amy Watkins of Purdue University, who was one of the original founders of the student-run pro-bono consulting group. Encouraged by Prof. Rod Little and Bhramar Mukherjee, Michael Elliott was the original faculty adviser, and Maria Larkin the first president. Our first project involved reviewing surveys and discussing sampling with Career Alliance, a non-profit workforce development organization in Flint, MI, to help them assess discrepancies between health care training opportunities and health care job openings in Genesee County. Since then STATCOM-UM has had eight student presidents, a faculty co-adviser (Cathie Spino), several faculty project advisers, and approximately 100 student volunteers from Biostatistics, Statistics, and Survey Methodology. We have worked on more than 60 projects for non-profits, primarily (but not exclusively) in the southeast Michigan area. Some projects are short, involving perhaps one or two meetings and discussions; others involve development of surveys or databases; and yet others involve full scale analysis and report writing.
Notable STATCOM Projects
- To determine the extent of psychosocial risk for waitlisted transplant patients
- To assess whether time on the transplant list impacts psychosocial risk factors
- How: logistic regression
- We found an in odds of worsening social support (OR=1.07, p = .04) and learning barriers (OR=1.16, p<.01) the longer candidates wait for transplant
- list of current and past partner locations
- 2015-2017 visitor data from PantryTrak
- Aim 1: determine optimal pantry locations for meeting food-assistance needs, prioritizing areas with low income and attempting to maximize the number of people served
- How: optimization model that weights households according to their need score
Statistical Analysis of Caseload Size and its Association with Clinician Efficiency at The Children’s Center of Detroit
- Assess associations between caseload and two measures of clinician efficiency: expected pay (how much each clinician earns for TCC each month from their appointments) and attendance rate.
- Mixed Effects Model
- For clinicians with caseloads less than 70, increasing a clinician’s caseload size leads to an increase in expected pay.
- For clinicians with caseloads greater than 70, increasing a clinician’s caseload size leads to a decrease in expected pay.
- Clinicians with higher attendance rates can handle a larger number of cases before their expected pay begins to drop.
- Caseload size and attendance rate are inversely associated. Although the two variables are significantly associated, the association is weak.
- Aim: To determine if significant differences existed in duration of the job search, salary, and loan debt by race/ethnicity and gender.
- Evaluated career outcome surveys for recent graduates of the School of Public Health
- PDF Summary