• Dec 28, 2016

Why Trial Error Matters for Your Data Quality

It’s crunch time for numbers as farmers plan for the 2017 growing season. But the plans you make are only as good as the data you’re analyzing. Trial error indicates whether your data is legitimate or not. And if your data is “iffy,” any resulting actions you might take will be equally suspect. Here’s a look at what trial error is and its role in producing quality data.

Trial error calculations helps us understand variation within replicated trials that cannot be accounted for by controlled treatments. In a nutshell, trial error comes from factors that we cannot see or anticipate that affect our outcomes.

For example, if you planted the same corn hybrid in two different locations, it may yield differently due to disparities in nutrient profile, soil moisture, pest pressure, etc. Now let’s say we implemented a trial design that accounted for all of these factors and created a scenario where both fields were, in theory, identical. What if they still yielded differently? It is this unexplained variation that cannot be easily accounted for that we call “trial error.”

How would you explain a veteran baseball player who averages 10 home runs a year suddenly hitting 50 homers one season? First, you’d look at the different variables. Are the baseballs wound tighter? Are the stadiums smaller? Are pitchers less capable? If everything remained the same, is it safe to assume that player will hit 50 home runs again next year? Probably not. But over time and repeated trials (seasons), we’ll be able to tell with certainty if the breakout season was an anomaly or a true indicator of performance. 

In short, replication and low trial error are necessary to be confident in your data.

Answer Plot® Program Trial Error
To give you an idea of how we’re able to deliver such quality data, with low trial error, let’s take a quick look at the numbers. In 2015, the WinField Answer Plot® Program tested 231 corn hybrids replicated 12 times at 191 locations across the country. We also collected a total of 5 million data points from our trials. This level of local, regional and national testing allows us to ensure the validity of our data, and more important, it allows us to deliver more predictable product recommendations such as seed prescriptions. 

In addition, the WinField United agronomy and product development team adheres to a strict set of standards that shepherd the communication of quality, reliable data. This set of data and technology principles mean that any piece of data published by WinField® United has been extensively vetted for accuracy and quality.

It’s crunch time for numbers as farmers plan for the 2017 growing season. But the plans you make are only as good as the data you’re analyzing. Trial error indicates whether your data is legitimate or not. And if your data is “iffy,” any resulting actions you might take will be equally suspect. Here’s a look at what trial error is and its role in producing quality data.