Finalizing Your On-Farm Trials
Palle Pedersen, Soybean Extension Agronomist
Originally published in Iowa Farm Bureau Spokesman, August 24, 2005

During the past year many farmers were asked to test new products such as varieties, inoculants, fungicides, insecticides, and seed treatments on their farm. Some products were even donated without any expectation of data from the company. It is always easy to accept products, but it is not very often the data is taken from these trials and used since it takes time. Why not take advantage of the product and see if it will work on your farm? Many believe that the visual effect is all what matter but we all know that isn't the case with soybean. Soybean yield is mostly determined by seed number and seed size, and who can see that?

There are some general guidelines you need to consider when you take data from your farm. The most important help I can give you is to be sure you do not fool yourself. If the experiment has not been laid out in the proper way the data may not help you at all. You want to be sure that the data is reliable and true. The success of your research trial will therefore depend initially on how well it is planned.

Many times much effort and expense is put into comparisons that are not valid. Certain statistical principles must be maintained. Otherwise, the information you obtain could be meaningless or, worse yet, deceptive. Replication, randomization, and use of a control are a must in conducting a trial since it helps us separate out treatment effects from natural levels of background variation. In this way, field variation and other factors will not discredit the reliability of the results. Treatment effects can be measured in many different ways. The most accurate way is to use a weight wagon, which can be rented from most elevators. Yield monitors can also be used but is not always that accurate.

Probably the biggest mistake to do is not to include replicated control strips in a field if you change a production practice. Most of us have an abundance of ideas for new production practices and products. Sometime those ideas can be implemented quickly and easily. At other times, a major investment is needed. Before adopting another management system or management practice, it is important to weigh the costs and benefits, but also consider the effect on long-term profitability and risk exposure. A friend of mine is a good example of this. He changed his entire tillage program on his farm to a more expensive program this year but did not leave strips with his previous tillage practices. I have been preaching to him so many times, but it seems that someone else has a better "connection" to him than me. I know that he will probably come back to me someday and complain about his profit margin on his soybean but what can I say?

It should be a standard practice to have on-farm research on your farm since it can be a powerful decision-making tool for you if you do it right. The reason to conduct on-farm research is to document whether a given product or production practice works on your farm to give you a better profitability. Many farmers believe that you have to be a farmer with at least 1000 acres and have a yield monitor on your combine to be able to do on-farm research. That is not the case. If you were not one of many that conducted on-farm research this year then think about it for 2006. You may have observed things in your field when you scouted this year that may be worthwhile investigating in the future. That specific problem could be the 30 bushels that you are missing.