On-farm research has reached the Henderson operation, where every crop input must now explain what it contributes before being invited back next season.
The policy was announced at the kitchen table. Fertilizer was asked for references, a biological product was asked for local data and a salesman requested that the review recognize his excellent relationship with the family.
“Does the relationship yield?” asked farmer Elaine Henderson.
The salesman adjusted his embroidered polo shirt and requested a less confrontational atmosphere.
On-Farm Research Introduces Performance Reviews
The serious model is farmer-led comparison under actual field conditions. Practical Farmers of Iowa’s research opportunities describe randomized, replicated strip trials, including work comparing nitrogen rates. The purpose is to measure outcomes, rather than award the most convincing story a production budget.
Henderson said the distinction arrived just in time. Her farm had accumulated several inputs whose primary measurable effect was ensuring the salesman remembered her birthday.
She began with a simple question: would a proposed treatment increase net returns on her ground?
Her father said the answer could be found by looking at the nicest corner of the field. Her agronomist observed that the nicest corner had unusually good soil, excellent drainage and thirty years of being chosen for photographs.
“The corner is talented,” Henderson admitted. “We may have been giving the jug too much credit.”
The Control Strip Refuses to Feel Guilty
The family laid out comparisons and kept records. Henderson’s father regarded an untreated strip as an expression of insufficient ambition. He drove past it twice daily to let it know the family expected better.
At a field meeting, the salesman presented a photograph of handsome corn. Henderson produced a photograph of equally handsome corn from another treatment. The exchange ended without either photograph explaining production costs.

A proper comparison, her agronomist said, needs care with field variability, replication, measurements and economics. One strip can suggest a question. It should not automatically settle an argument for an entire county.
Henderson liked that approach because farm decisions already contain enough things that cannot be controlled, including weather, prices and her father’s opinion of the planter settings.
The new system also allows a practice to earn its place even if it produces the same yield at lower cost. Her father initially objected that this made the yield contest harder to discuss at the café.
“Tell them the checking account won,” she suggested.
The Free Hat Enters the Evidence
The salesman offered a cap. Henderson accepted it as a cap. This was a major change from the previous system, under which promotional clothing had occasionally been treated as a minor agronomic credential.
The family’s informal survey found strong support for measuring results, provided nobody measured how long the machinery had been parked in front of the shop.
Henderson now recommends starting with a clear question and asking an Extension adviser or experienced research group to help design a useful comparison. A badly planned test can produce a remarkably confident answer to the wrong question.
The salesman has agreed to return with evidence. He will bring local results, total costs and a proposal that can be evaluated.
His polo shirt will remain welcome.
“It gets a seat at the table,” Henderson said. “It just doesn’t get its own treatment replication.”
Satirical disclaimer: No polo shirt was counted as a replicated trial. Our ceremonial human collaboration pairs two sentient beings—the world’s oldest tenured professor and a philosophy major turned dairy farmer—who remain available to argue about the underlying evidence.
