Progress 09/15/24 to 09/14/25

Outputs
Target Audience:During this reporting period, our primary outreach occurred through the development, submission, and dissemination of manuscripts that advance statistical and computational methods for turfgrass cultivar evaluation. The audiences reached reflect both scientific and applied communities. Academic Researchers. We engaged turfgrass scientists and plant breeders at universities who rely on NTEP data for cultivar evaluation. Manuscripts on latent scale models and Bayesian approaches to seasonality introduced new ways to address rater subjectivity, spatial confounding, and ordinal scale limitations, providing these researchers with reproducible frameworks to improve trial accuracy. Quantitative Scientists. Our work reached statisticians and data modelers interested in applying Bayesian hierarchical models, Gaussian Processes, and Item Response Theory to agricultural data. Manuscripts and presentations connected turfgrass research to broader communities in statistics, biometrics, and data science. Breeders and Applied Users. Commercial breeders and practitioners use NTEP results to guide cultivar release and adoption. By improving the fairness and comparability of ratings across sites and years, our manuscripts offered practical tools to accelerate breeding decisions and cultivar recommendations. NTEP Cooperators and Governing Bodies. Findings were shared with cooperators and advisory board members representing universities, seed companies, sod producers, and industry associations. Their feedback ensured that manuscripts addressed questions of direct relevance to ongoing trials. Extension Specialists. Extension educators, who translate research for professional turf managers and the public, represent a key secondary audience. Improved models enhance their ability to explain cultivar performance and sustainability outcomes. Students and Early-Career Scholars. Graduate students and early-career scientists were involved as co-authors and analysts. Manuscript work provided training in advanced modeling and data visualization, building capacity for future agricultural data science leadership. Summary. Overall, our manuscript activities reached seven main audiences: (1) academic researchers, (2) quantitative scientists, (3) breeders and applied users, (4) NTEP cooperators and governing bodies, (5) extension specialists, and (6) students. Through manuscripts, presentations, and collaborative discussions, we ensured that innovations in latent scale modeling, Bayesian seasonality analysis, and turfgrass database design were effectively communicated to those best positioned to apply them in research, breeding, and education Changes/Problems:During this reporting period, no major changes in project scope, protocols, or data management were required. The project remains aligned with the approved objectives and data management plan, and there were no changes involving animals, human subjects, or biohazards. However, three areas of note merit reporting: 1. Timeline Adjustments. Validation of the latent scale and Bayesian seasonality models took longer than initially projected due to the complexity of data processing and the need for additional computational resources. These adjustments did not alter the project's objectives but slightly delayed the transition to warm-season trial datasets. As a result, the majority of the reporting period was devoted to refining the models with cool-season turfgrass data. Work with warm-season grasses will be prioritized in the upcoming reporting period. 2. Data Availability. We encountered delays in obtaining complete and standardized metadata for some trial sites. In several cases, historical trial data required additional cleaning or transformation before being usable for model validation. While this slowed progress, it also underscored the importance of developing robust pipelines for handling inconsistencies and confirmed the value of our focus on data quality assurance. 3. Unexpected Outcomes. Although the primary focus has been on reducing rater subjectivity and modeling spatial and seasonal variation, the work revealed additional opportunities to detect rater-specific behavior patterns that may inform training or calibration of future NTEP cooperators. This outcome was not originally anticipated but represents a promising extension of the model's utility. Summary. Overall, no changes to the approved data management plan, scope, or compliance protocols were required. The main adjustment has been a shift in timeline, with warm-season validation deferred to the next reporting period due to the extended time required for model testing and data preparation. These changes are expected to have minimal impact on overall project goals and will be addressed through prioritized efforts in the coming year. What opportunities for training and professional development has the project provided?During this reporting period, the project provided training and professional development opportunities for postdoctoral researchers, and early-career scientists through direct involvement in data analysis, manuscript preparation, and model development. Participants gained hands-on experience with advanced statistical techniques such as Bayesian hierarchical modeling, Gaussian Processes, and Item Response Theory, along with skills in data cleaning, visualization, and reproducibility. Co-authorship on manuscripts and participation in peer review processes strengthened scientific writing and communication skills, while collaboration with senior researchers, NTEP cooperators, and industry stakeholders expanded professional networks and fostered mentorship. Collectively, these activities enhanced both technical expertise and professional readiness for future leadership in turfgrass science and agricultural data analytics. How have the results been disseminated to communities of interest?During this reporting period, results were disseminated to communities of interest primarily through peer-reviewed manuscripts, conference presentations, and advisory board updates. Manuscripts describing latent scale models, Bayesian seasonality approaches, and database innovations were ready to be submitted to international journals, ensuring broad visibility among turfgrass scientists, statisticians, and agricultural data researchers. Findings were also shared at professional meetings and workshops, where interdisciplinary audiences engaged with new analytical frameworks for cultivar evaluation. In addition, results were communicated directly to NTEP cooperators, breeders, and governing board members, providing practical insights into how the methods improve fairness, consistency, and comparability of trial outcomes. Drafts and preprints were circulated among collaborators and extension specialists to support timely application in breeding and extension programs. Through these channels, the project ensured that results reached both scientific and applied communities most likely to benefit from improved turfgrass evaluation methods What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period, we plan to extend our efforts by gathering data from warm-season turfgrass trials and using these datasets to further validate and refine the modeling framework. This will allow us to test the robustness of the latent scale and Bayesian seasonality models across different species and environmental conditions, ensuring that the methods are broadly applicable beyond cool-season grasses. We also plan to present results at additional scientific conferences and professional meetings, providing opportunities to share progress with turfgrass researchers, breeders, and data science communities, while obtaining valuable feedback to guide further model improvements. These activities will advance both the technical goals of model validation and the broader goal of disseminating innovations to the communities of interest.

Impacts
What was accomplished under these goals? 1) , 2) are accomplished completely. 3) and 4) are accomplished partially.

Publications