| Demonstration, Research — Each Serves Unique Purpose |
|
||||
| Ag Home Page |
| Agricultural Staff |
| Consulting Program |
| Plant Image Gallery |
| Internships |
| Recommended External Links |
| Search: |
| or Search by Keyword: |
 Pasture & Range: September 2002 Other Pasture & Range Articles |
|
by Bryan Unruh |
The Agricultural Division has a history of seeking answers to key questions that will lead to significant changes and improvements in existing practices and then transferring this information and technology to our clientele.
Over the years, we have conducted many demonstrations and applied research trials on our Foundation farms and ranches in southern Oklahoma. I would like to discuss how demonstration differs from research, what they are, and the purpose for each. The terms demonstration and research have been used interchangeably, but they are not the same thing. Now, I'm not here to be a nitpicker for terminology, but the two are different in their design and the results they can provide. I'm hoping you'll be interested in knowing a little more about how we do things here at the Noble Foundation.
The major distinction between research and demonstration relates to whether the objective is to learn something new, as with research, or whether it is to "demonstrate" something we've already learned. Demonstrations are designed primarily to show something. Examples of demonstration might be to show what a plant species looks like for identification purposes, or to show a concept, technique, system or process. In all of those cases, there is very little regard for gathering new information or comparing results. With a demonstration, we want to be able to show our clientele something already established or proven so they can learn something new. Conversely, then, research involves gathering new information. Let me hasten to say that a properly designed research project can do both — that is, we can gather new information and show, or demonstrate, various techniques, systems, etc.
At this point, you may be asking why a demonstration can't be used to gather new information or answer a particular question. For example, suppose an advertisement claims that a new rye variety, "X," will produce more total forage than the old standard variety "A." We decide to test this claim using a field demonstration. We plant side-by-side plots of equal size, one to variety X and the other to variety A, and measure the total forage produced from each plot. At the end of this exercise we determine that the greater yield was measured from the plot planted to variety X. Shouldn't we declare that variety X yields more than variety A?
Making this conclusion presumes that the yield difference between the two plots was due only to the forage yield character of the two rye varieties and nothing else. The error of this conclusion is that the yield from these two plots will always be affected by many other factors. This will include a host of factors related to soil variability, including texture, moisture, slope, fertility, or any other factor(s) that could affect one plot more than the other (e.g., insect damage). In fact, if we plant the two plots to the same variety and carefully measured forage yield, the yield from the two plots would not be equal. A characteristic of all experimental material is variation, particularly when studying biological systems.
From what we have discussed, it may seem that we have no hope of collecting information to determine if rye variety X yields more than the old standard. However, we have actually uncovered a key to how we conduct research! The key is this: if we can design an experiment in such a way that we can measure this variability, we can separate it from the answer we seek. Another key, though it may not be as apparent, is that we design the experiment so we get an unbiased estimate of the treatment effects (i.e., a good estimate of forage yield). In all experimental designs, the two tools we use to help us accomplish this are replication and randomization.
This is an introduction to the experimental methods that must be employed to conduct research. In my next article, I will discuss why we must use these and other statistical design tools to set up experiments and find answers to our questions.