StatsPD@Waite meeting: In-plot sampling in field experiments – discussion forum

Every month, the professional development meetings of statisticians and data scientists at Waite, known as StatsPD@Waite, bring together specialists in various aspects of data sciences in agriculture from Waite, Roseworthy and Adelaide.

Date: Tue, 14 October 2025, 10:00 – 12:00 (Adelaide)

Location: Online via Zoom

Contact: biometryhub@adelaide.edu.au

In-plot sampling in field experiments – discussion forum

In-plot sampling is central to experimental work in agronomy and plant and soil sciences, yet its statistical implications are often overlooked compared with the design of treatment allocation. While field experiments routinely use randomized complete block designs and related structures, the selection of plants, soil cores, or other units within plots is frequently based on convention rather than statistical reasoning. Practices such as diagonal transects, “representative” judgment, or ad-hoc subsampling can affect efficiency, bias, and the interpretation of treatment effects, yet these effects are rarely discussed.

This facilitated discussion with Dr Ray Correll will bring together data analysts, statisticians, and agricultural scientists to examine better approaches to in-plot sampling. Rather than promoting a fixed methodology, the forum will explore assumptions, trade-offs, and opportunities associated with different strategies. Key topics include balancing random and systematic coverage, handling destructive versus non-destructive measures, designing subsampling for multiple responses, and accounting for spatial or temporal variation within plots. We will also consider how in-plot data can be aligned with plot-level analysis and how discrepancies between common agronomic practice and statistical principles can be addressed.

The goal is to stimulate dialogue and build towards identifying priority areas for guidance and collaboration in developing robust, practical sampling strategies.