Measurement & Experimental Science

Representative Sampling, Replication & Controls

Reliable cultivation comparisons require representative sampling and enough replication to separate a treatment effect from natural plant-to-plant variation.

Key concepts

  • A single plant can be atypical
  • Randomization reduces systematic placement bias
  • A control provides a reference condition
  • Replication improves confidence but does not fix a biased design

What to measure or observe

  • Define the experimental unit before starting
  • Randomize or balance positions when possible
  • Measure the same traits with the same method
  • Record excluded or failed samples transparently

Common mistakes

  • Treating multiple leaves on one plant as multiple independent plants
  • Comparing different cultivars as if genetics were controlled
  • Changing several variables at once

Visuals this lesson still needs

  • Replication diagram
  • Randomized layout example
  • Control-versus-treatment design
  • Pseudoreplication warning graphic