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

Don't Get hung Up on One or Two Bushel Differences

Don't Get hung Up on One or Two Bushel Differences
Even in replicated trials it's virtually the same.

If you're looking through yield trial books or looking at plot results form strip trial, remember that data that's not replicated in some way, either repeated several times at one location or repeated in the exact same design at several locations, you're just looking at numbers, not data. The trend may seem interesting, but there's no way to know that if the same trial was run again, the same results could be expected.

Likewise, even in replicated trials, if the yields of two treatments aren't significantly different, even if one is a couple bushels better than the other, you have to assume that there is no difference between them. You could run the same trial again and the winner could be reversed by the same amount.

One to two bushels difference per acre even in a replicated trial where the two aren't significantly different is basically saying they're the same, notes Bob Nielsen, Purdue University corn specialist. The fact that one out yielded the other and is listed above the other may plant a seed in your head that it is better, but science says otherwise.

Scientists live by what's called LSD, or least significant difference. If you run a replicated experiment and hybrid A yields 196 bushels per acre and hybrid B yields 200 bushels per acre, but it takes 5 bushels spread to make it significantly different, than for all practical purposes both are considered the same.

Four bushels may seem like quite a bit, but in an experiment where there are many variables, including soil type and human error, and perhaps weather differences in bigger plots, plus PH and soil fertility variations, that much difference could happen by chance, Nielsen says.

Suppose the yields are 194 and 200, at a 5 bushels spread to make it significantly different. Now you can say that hybrid A is not as good as hybrid B. If you were at the 0.10 LSD level, that means you would be 90% sure that there was a real difference between the two hybrids or treatments, and that it was not due just to chance.

So as you look at yield results this fall, either for hybrids or agronomic practices, keep in mid that even if one out yields the other, it's whether it's significantly different or not that matters.

Hide comments
account-default-image

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.
Publish