One thing Dave Nanda always encourages is getting data from as many sources as possible before you make a decision on trying a new hybrid, variety, or farm practice. That typically means seeking unbiased results.
If it's on hybrids and varieties, you can check university plot results. For example, Purdue University and their Plot trial program now posts data from a five-state area on their Web site. It includes data from neighboring universities that run similar hybrid and soybean variety test programs. The advantage, believes Nanda, a plant breeder, crops consultant and now director of genetics and technology for Seed Consultants, Inc., is that you get information from various environments.
The ideal situation would be to test practices on your home farm. If you do, make sure that you replicate them and follow good testing procedure to eliminate experimental error. Bob Nielsen and Shaun Casteel, Purdue University Extension agronomists, are emphasizing the value of replicated on-farm trials which are set up correctly to account for experimental error.
When you're checking out results from somewhere other than your farm where you weren't involved in setting up the test, either read the fine print if it's a written report or call and ask questions. Was the plot replicated? Did one practice get some kind of unfair advantage? Were spoil types across the field similar? How many times was the experiment repeated?
Getting answers to all of these questions will help you decide how much faith and trust you can put in the results form that test as you factor it into your decision-making process, Nanda notes.
Nanda is also one who believes in trying something instead of jumping into it whole-hog. That works well for hybrids and varieties. It may not work so well if you have to buy a tool, like a vertical tillage tool, to make the switch to a different practice, or if you have to equip your planter with starter fertilizer attachments to compare starter vs. no starter on your farm.
In the latter two examples, finding information from others who have run these trials would be helpful. Starter fertilizer data should be readily available. Note planting date, weather condition, soil fertility levels and other factors that could affect whether or not there might be a yield pay off.