Agricultural Statistical Consulting Center | ASCC
“All models are wrong, but some are useful.” - George Box (1976)
Mission
The Agricultural Statistical Consulting Center (ASCC) supports the ACES community by providing expert guidance in experimental design and data analysis using sound, innovative, and appropriate statistical methods. We serve as a reliable and accessible partner, working collaboratively with researchers and students to strengthen the rigor of scientific inquiry and ensure that research questions and hypotheses are addressed objectively. Through these partnerships, we help translate scientific findings into improved and more sustainable agricultural practices. Our work supports the efforts of scientists in the college to reduce environmental impacts, enhance food quality and security, and promote agricultural systems that can meet global needs responsibly. We are committed to advancing knowledge and fostering a better, more sustainable future for current and future generations.
Scope
The ASCC provides support in scientific data-driven research, including the design and statistical analysis of experimental and observational data, to advance research and discovery within the ACES community and its collaborative partners. Our team of experienced data scientists builds and adapts models to extract insights from data, addressing questions related to inference and prediction within both statistical modeling and machine learning frameworks. The Center assists research partners at all stages of their work, from experimental planning and setup to data collection, analysis, and interpretation. Collaborators are encouraged to include ASCC statisticians on grants, reflecting our role as an integral partner in promoting rigorous and reproducible research.
Here to Help
The Agricultural Statistical Consulting Center (ASCC) provides a wide range of statistical services to support research in agricultural and biological sciences. In addition to the services listed below, we offer consulting on a variety of other statistical approaches tailored to specific research needs.
Types of assistance provided to scientists in ACES:
- Design of experiments and their analyses
- Analysis of observational data
- Grant proposals: designing and recommending appropriate statistical procedures for data collection, analysis, and validation
- Writing clear and rigorous statistical sections for scientific reports and manuscripts
- Statistical analyses for scientific manuscripts
- Summarizing results through effective tables, graphics, and figures
- Training on statistical methods for non-statisticians
Frequent topic areas:
- Linear and non-linear regression, including time-varying regression coefficients
- Bayesian and classical Analysis of Variance (BANOVA and ANOVA) for crossed and nested experimental designs
- Fixed, random, and mixed models
- Modeling of ordinal, nominal, and count data
- Non-parametric regression
- Multivariate analysis
- Variable selection in regression models
- Spatial statistics
Get Started
Consulting services are currently offered by appointment and certain walk-in hours. Please contact ASCC via email (listed next to personnel) to set up an appointment during the academic year.
Walk In Hours
Talk to An Expert
Ciro Velasco-Cruz
 |
Title: Assistant Professor |
Walk-in service: Wednesday and Thursday from 10:30 to 12:30.
Dr. Velasco-Cruz is an experienced statistician specialized in classical and Bayesian methods, with extensive expertise in statistical consulting, research collaborations, and teaching. His current research focuses on modeling biological and agricultural data using advanced statistical and data science techniques.
His methodological expertise includes:
- Spatial statistics
- Dynamic linear and non-linear models
- Non-parametric Bayesian analyses
- Bayesian variable selection
- Supervised and unsupervised learning and classification
- Graphical models
- Dimension reduction techniques
These approaches are applied to a wide variety of data types. For example, spatial and dynamic models handle multi-location time-series data; variable selection identifies key predictors for inference or prediction; dimension reduction simplifies high-dimensional genomics or biological data; and latent variable models are ideal for classifying outcomes such as disease progression or product quality.
Dr. Velasco-Cruz integrates statistical modeling and machine learning with domain expertise to build robust, insightful models that enhance understanding, uncover new insights, and support reproducible research. He routinely uses SAS, R, MATLAB, and SAS-JMP, and combines R with C++ for large or customized analyses to achieve maximum efficiency.
To schedule an appointment with Dr. Velasco-Cruz, email him at: cvelasco@nmsu.edu