Below you will find a list of preceptorship projects that faculty have offered as potential project ideas. Note that this list is not exhaustive and you may approach faculty members for other project ideas.  Also, please be reminded that a 1 s.h. preceptorship option is available for students who already have sufficient experience in biostatistical collaborations (as determined by the student’s advisor and the Director of Graduate Studies). In this case, the student would generally prepare a preceptorship closely related to work they have conducted for their research assistantship (or some other collaborative research endeavor). The hours spent on the preceptorship must be in addition to any work the student spends on their regular paid assistantship (e.g., work as a research assistant). 

Preceptorship Projects Available for 2025-26

(revised February 4, 2026)

1. Potential for causal inference in diabetes clinical trials (Emily Roberts)
   This project would explore some aspect of integrating causal techniques including counterfactuals and propensity scores into T1D studies for purposes of integrating external datasets or surrogate markers.
2. Potential use of mediation methods with cancer registry data (Emily Roberts)
   This project would explore hypotheses of mediating pathways in cancer survival and sensitivity analyses of unmeasured confounding variables.
3. Measures of diversity and equity (Jeff Dawson)
   Dr. Dawson is interested in supervising a preceptorship focusing on statistical methods in justice, equity, diversity, and inclusion (JEDI). This project would investigate existing and novel metrics of diversity, representativeness, and/or income disparities, through simulations and analysis of publicly-available data. The project will extend recent work to be presented by JEDI experts at the Joint Statistical Meetings in August of 2023, where Dr. Dawson was an official discussant.
4. An Empirical Study of Post-Selection Inference in Biomedical Journals (Ryan Peterson)
   Many biomedical studies use variable selection methods before reporting p-values or confidence intervals, and often do not account for the uncertainty introduced by the selection step. In this project, the student will develop a reproducible R-based workflow to sample biomedical journal articles and extract statistical methods from full text using a combination of rule-based text analysis and AI-assisted review. Automated and assessments using large-language models (LLMs) will be validated through manual review of a subset of papers to estimate how often unadjusted post-selection inference occurs and to evaluate the reliability of LLMs as screening tools for identifying statistical issues.
5. Four potential projects with Kai Wang
   Please contact Dr. Wang to discuss the data use.
   5.1 Summary statistics Mendelian randomization analysis of tinnitus and hearing difficulty in noise.
   5.2 Improving polygenic scores using deep learning.
   5.3 Prediction of PCB contamination level using deep learning.
   5.4 Survival analysis of risk factors on conversion time to glaucoma.
6. Breheny project: Improving tests to detect mRNA-miRNA binding sites: Modern deep sequencing methods have enabled researchers at the University of Iowa to collect genome-wide data on binding interactions between messenger RNA (mRNA) and micro RNAs (miRNA). When such an interaction is present, it leads to a build-up on reads in a particular location on the genome. The current way that these investigators test to see if such a “bump” is real, however, treats each nucleotide independently. This is flawed, however, since reads typically span many nucleotides and are thus spatially correlated. The goal of this preceptorship is to test for these interactions in a more sophisticated manner and cut down on false positives.
7. Multiple Structural Breaks Detection through Genetic Algorithm (Gideon Zamba)
   Under stimuli and workloads, the human body tends to display discomforts and covert proximal cognitions that can manifest through physiological responses. One such response can be in a form of skin sweats, easily capturable via wearable sensors. The sensors capturing electrodermal activities (EDA) record big data serially (4Hz, 8Hz), per subject, over a duration of an experiment (approximately 35 min). This project aims to study serial EDA data collected on subjects from an experiment being conducted in neuropsychology at the University of Iowa, with the goal to detect structural breaks corresponding to epochs of learning activities and assess the role that biofeedback plays in efforts to engage people in a learner space. The project mobilizes technological innovations in neuroimaging (fNIRS), wearable sensors monitoring covert cognitive activity, monitoring arousal states under workload, and video data outputs to address learners’ emotional discomfort. The project will focus on structural breaks detection though a genetic algorithm stochastic search across the spectrum of EDA data.
8. Burnout Among Servicemen: A case of the Russian-Ukraine War (Gideon Zamba)
   Studies of war veterans estimated that 95% of military members consider burnout to be the leading cause of separation, retirement, and interpersonal disorders. Burnout, defined as a syndrome of emotional exhaustion and feelings of workplace failure that occurs in response to chronic exposure to occupational stressors, if not attended to and properly addressed, can trigger, or induce emotional disorders, feeling of discouragement, frustrations, worthlessness, and depression among servicemen. The current project involves researchers from the University of Iowa and investigators in the Psychology Department at Tara Shevchenko University in Kiev, Ukraine. The data were gathered on Ukraine’s servicemen at the frontline of the Russian-Ukraine war (n = 400). Burnout Assessment Tools (BAT), Interpersonal Guilt Rating Scale Self Report (IGRS-SR), Basic Psychological Need Satisfaction and Frustration Scale measurements (BPNSFS), will be cross-studied as functions of servicemen socio-demographic characteristics such as: gender, age, education, marital status, number of children, combat operation and other characteristics. Specific hypotheses will be studied and tested, and recommendations will be made with direct policy implications regarding burnouts among frontline servicemen.
9. Cancer research project (Grant Brown)
   It has been widely reported that cancer rates in Iowa have been either increasing or failing to improve at the rate expected relative the rest of the USA over the past several years. This is a challenging problem from a statistical/epidemiological level, because “cancer” refers to a diverse class of conditions, and arises based on a lifetime of exposures and predispositions. In addition, the data available to researchers on cancer incidence and exposures are imperfect and often highly aggregated. In this project, we’ll attempt to build on current work using spatial models to understand the distribution of cancer burden within and beyond the state of Iowa in order to pursue explicitly cumulative models of exposure risk. A particular focus will be on radon and lung cancer, and we will conduct the work in a Bayesian context.  
10. An Investigation of the role of GLP-1 Receptors in Vision Loss: The Case of NAION patients (Gideon Zamba)
    Glucagon-like peptide-1 (GLP-1) receptor agonists are a class of anorectic drugs that help reduce blood sugar and energy intake by activating the GLP-1 receptor. GLP-1 receptor agonists are being used for weight loss and diabetes and are becoming the most widely used medications for these conditions. Recent research provided some evidence of an increased risk of non-arteritic anterior ischemic optic neuropathy (NAION) in patients using GLP-1 receptor agonists. NAION is a condition that is easily associated with the risk of a permanent vision loss. However, clinicians are not settled on the underlying causal mechanism of NAION. Although there have been suggestions that NAION conditions are linked to abrupt decrease in blood flow toward the optic nerve, there appears to be a paucity of research work in support of this hypothesis. Is there any viable association between NAION and GLP-1 receptor agonists? Are subjects under GLP-1 more likely to develop visual loss? The current research study will embark on an exploratory retrospective chart review of the record of n = 400 NAION patients, their pattern of vision loss if ever, their visual imaging data, their pattern of  GLP-1 receptor agonists usage or lack thereof, their health history, their diabetics and obesity status, and other prognostic factors that can contribute toward elucidating on time to visual loss in both with or without GLP-1 receptor agonists usage. The project will mix time to event modeling and image analysis to explore this potential association amidst a host of prognostic factors.
11. A Quality Adjustment Survival Tool for a Patient-centered Personalized Care in Oncology (Gideon Zamba)
    Cancer patients outside of randomized controlled trials are increasingly being presented with numerous and complex treatment choices that may have serious implications on their health outcomes and on their quality of life (QOL). These treatments can differ substantially in toxicity (serious adverse reactions, including the risk of second primary malignancies) and in efficacy outcomes of direct relevance to the patient such as survival and hazard rates. Quality-adjusted survival (QAS) analysis is used to ascertain if, for example, potential prolongation of life outweighs the risk of diminished QOL from a patient’s perspective. Recent developments in precision medicine and personalized care advocate for incorporating patient-specific benefits and inputs into medical decision-making rather than the usual ‘regressing through the mean’, i.e., acting on what works for the ‘majority’ in the population—as the patients and their families are the ones that have to live with the consequences of a treatment versus an alternative. Quality-adjusted time without symptoms of disease and toxicity (Q-TWiST) is a useful statistical method that incorporates progression, survival, toxicity, and utility measures into a single metric to provide an integrated measure of clinical benefit. However, implementing Q-TWiST relies on complicated statistical modeling applied to patient-level data which makes its direct clinical application and its use in decision-making challenging. This project proposes to develop a QAS effect size (QASES), a measure that incorporates both overall survival (OS) and toxicity-free survival (TOX) to interrogate for the patient the level of a severe or serious toxicity that is worth the potential increase in survival. Given published data or summary information from a completed RCT with time-to-event endpoints, we will design a conversion/standardization approach in which the OS and TOX log hazards ratios are each converted to a standardized mean difference using variance-stabilizing tools. When there appears to be a violation of proportional hazard assumptions in a trial’s results, the proposal will deploy restricted mean survival time tools for building a simple and unified QASES for decision-making. This project will involve some theoretical manipulation, some simulations, and a future R package development for dissemination.  The project is run in conjunction with statistical researchers at the Mayo Clinic-Rochester.
12. Statistical advice from LLMs: How bad could it be? (Dan Sewell and Ryan Peterson)
    Anecdotally, physician researchers are increasingly turning to large language models (LLMs) to determine how to set up their study, analyze their data, and interpret their results.  Drs. Sewell and Peterson are interested in systematically exploring how various LLMs respond to such tasks, what errors occur, including LLM-specific error rates and which errors are consistent across LLMs.
13. Anterior Cruciate Ligament Re-Tear (Gideon Zamba)
    The anterior cruciate ligament (ACL) is a key intra-articular structure of the knee joint, oriented obliquely to connect the femur to the tibia. Its primary biomechanical function is to resist anterior translation of the tibia relative to the femur and to contribute significantly to rotational stability of the knee. ACL injuries are prevalent among athletes participating in cutting and pivoting sports such as basketball, football, soccer, and lacrosse. The incidence of ACL injuries among adolescent athletes is steadily increasing, approaching nearly 1 per 10,000 AEs for female athletes, who were almost 1.5 times as likely as male athletes to suffer an ACL injury across all adolescent sports. Clinical management of ACL rupture frequently involves surgical reconstruction, most commonly utilizing autografts such as bone–patellar tendon–bone, quadriceps tendon, or hamstring tendon. The procedure aims to restore knee stability and function, with graft selection influenced by patient-specific factors, activity level, and surgeon preference. Even though 83% of patients return to their pre-injury level of activity, re-tear rates have been documented to be as high as 30%. Contributing factors, such as graft type, surgical technique, time between surgery and release to RTS, and age have been gaining more attention in the orthopedic surgery literature. Functional testing, level of preinjury activity, quadriceps and hamstring strength, and psychological testing have been shown to aid in determination for RTS. Researchers at the University of Iowa Department of Rehabilitation Therapies and Sports Medicine are interested in the mechanism by which graft type, quad strength, rate of force development, time to recovery, and other risk factors relate to ACL re-tear rates among athletes. Statistical learning tools will be deployed for this exploration.