Farm Progress is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Serving: IN

Why You Should Care About Statistics in Agriculture

Why You Should Care About Statistics in Agriculture
Here's how to tell if you can rely on test plot data to make decisions.

If you talked about LSDs and statistics at a farmer meeting even 10 years ago, eyes would begin to glaze over in about 30 seconds. Maybe they still do for some, but farmers who have perked up their ears about data and what it can do know that having reliable data is important.

Jeff Ezell, a seed rep with Winfield, says that you need a handle on whether you can rely on data or not, and how much trust you can put into it. Statistics is the tool that lets you do that.

Why statistics matter: Jeff Ezell says you can have more confidence in data from trials which were replicated, especially if they were replicated many times.

The big term you need to understand is 'least significant difference.' In layman's terms it's a measure of how much experimental error is in the data. The more replications at the same location or the more the number of locations with the same plot, the more likely that the data will contain less error.

What LSD does is estimate how likely it would be to repeat the experiment and get the same results. Were the differences in hybrids, varieties or practices due to real differences in the product or method, or could it just have been by chance?

Related: Solar Panels, Field Data Inspire Farmer's Corn Yield Trials

Sometimes you will see LSD's on 10 to 20 at the bottom of a table if you read the fine print. If a strip trial wasn't replicated and statistics weren't applied, you won't find an LSD.

If you look at data from universities and some companies that replicate trials, you should find an LSD.

This example shows what it means, he says. Suppose the LSD is 3. That's very low and can be achieved by lots of replication. Some of his company's plot data is that precise because they run the same replicated trial at many locations, he says.

If the LSD is 3 and hybrid and the top three hybrids are 200, 199 and 198 bushels per acre, all have the same chance of being equally as good. However, hybrid D at 190 bushels per acre is significantly lower yielding. You can have confidence that it will yield less compared to the other three if you plant it.

However, if you don't have as many replications and there is more error in your experiment, and the LSD is 20, you can't say for sure with the same set of numbers if the first three hybrids are truly better than hybrid D or not.

In fact, if hybrid E yielded 181 bushels per acre, you still don't know beyond a reasonable doubt if it is lower yielding. You would conclude that there is no significant difference in yield between all top five hybrids.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.