Field Test(redirected from field-testing)
Also found in: Dictionary, Thesaurus, Medical.
field test[′fēld ‚test]
an experiment conducted under field conditions that approximate production conditions to establish the relationship of the quantity and quality of the harvest to the farming conditions and technology; a form of agronomical research.
A field test is organized according to a definite methodology that specifies a number of variations and replications; the area, shape, direction, and arrangement of the plots; and methods for keeping a record of the harvest. The results of the field tests serve as the basis for the widespread introduction of new farming procedures and plant varieties into agricultural production.
A field test should be made under conditions typical for agricultural production, that is, on a soil typical of the specific region, in the same system of crop rotation, and with procedures of the same level of sophistication. A field test is conducted according to a definite plan that consists of a limited number of variations differing only by the one factor (a certain procedure or plant variety) being tested. The number of plots and variations is determined by the number of replications used in the test. Replication makes it impossible to increase the accuracy of the test results and to estimate the degree of reliability of differences obtained in the test between the yields of the variations being compared. For reliable results, a field test should have at least four cycles of replication, and in certain cases, when a higher degree of test accuracy is required, six or eight.
The field used for the test should be uniform in relief and soil variety; the same tillage, fertilization, and crop rotation should be maintained for three or four years before testing. The size of the field depends on the size of the plots and on the number of variations and replications used in the test. The plot size and the number of replications are determined by characteristics of the crop, the subject of the test, the relief, variations in the soil cover, the implements and machines used during testing, and the requirements for test accuracy. The size of a plot is usually the minimum area that will, under the given conditions, ensure the necessary accuracy and make it possible to do all field jobs, including recording the harvest, with maximum mechanization. In practical test work, square and retangular plots with areas of 50-200 sq m (sometimes 300 sq m and more) are most commonly used. The plots are arranged in one row or several rows with their long sides touching. (The arrangement depends on the relief and configuration of the field.) Where the plots are arranged in several rows there are one, two, three or more replications, each replication being within a row. The arrangement of variations in the plots within a replication may be designed in a certain order identical for all the plots; it may also be random, or standard. Methods based on the random arrangement of variations are most important.
Protective belts are planted along the edges of the plots and field before testing. The harvest from them is gathered separately. The protective belts are necessary to eliminate error introduced by adjacent plots (for example, a different fertilizer). In addition to recording the harvest, a field test includes observation of plant growth and development, soil condition, and meteorological conditions. The content of the program varies depending on the purpose of the field test.
The yield from a plot is often determined by recording the yield of the entire plot. In tests with cereal, fiber, and herbaceous crops, the yield is also recorded by sample sheafs. Yield data are statistically processed, which makes it possible to establish the degree of precision of the test and to show that differences obtained in the test by comparing the average yields of different variations are reliable, that is, they significantly exceed the magnitude of random error. If the figures are within the error limits, they are not reliable. One of the most common methods of statistical processing of field test data is analysis of variance, which makes it possible to find the general error of average yields for the test as a whole and one general error of difference for the average yields of any pair of variations within the test that are compared.
REFERENCESKonstantinov, P. N. Osnovy sel’skokhoziaistvennogo opytnogo dela. Moscow, 1952.
Vol’f, V. G. Statisticheskaia obrabotka opytnykh dannykh. Moscow, 1966.
Dospekhov, B. A. Metodika polevogo opyta, 3rd ed. Moscow, 1973.
F. A. IUDIN