Calibration of Soil Tests
Although the chemical analysis of soils presents some difficulties, perhaps the greatest problem in a testing programme is the calibration of the tests. It is essential that the results of the soil tests be calibrated against crop responses from application of nutrient in question. This in[formation is obtained from field and green house fertility experiments is obtained from field and green house fertility experiments conducted over a wide range of soils. Yield responses from the rate of applied nutrients can be than related to the quantity of available nutrients in the soil.
It is usual in experiments on calibration of soil to use a group of soils which range in fertility level from low to high in respect of particular nutrient under study. A crop is usually grown on these soils with varying doses of nutrient under study over a basal dose of other nutrients in adequate amounts. Data on growth yield and uptake of nutrients are recorded. The calibration may be accomplished in several ways. The most common is to prepare the scatter diagram by plotting soil test values against percentage yield on a graph and calculation of correlation coefficient between soil test values and per cent yield response as crop yield.
Crop yield with adequate nutrient yield of control
_______________________________________ X 100
Crop yield with adequate nutrient
The correlation coefficient (r) is an index showing the degree of linear relationship between two variables. The coefficient of determination (r2 %100) is a measure of percentage variation in one variable that is associated with variation in other variable. By correlating a large number of percentage yield and soil test values a value of ‘r’
is obtained which is statistically significant shows that the relationship is not due to chance.