ERROR OCCURRENCE PROBABILITY OF TYPE I AND II IN MONITORING OF A SEEDER-FERTILIZER

The monitoring of the seeder-fertilizer performance throughout the sowing grains becomes essential to ensure its operation and to determine in which moment the pause intervention during the operation should occur. However, a way to analyze the performance of the seeder-fertilizer can be done through the individual values control cards, which detect the presence of eventual causes due the seeding, becoming an important analysis/manager tool. In this way, this paper focuses in evaluate the probability of occurrence of the errors type I and II in the operational performance analysis of a seeder-fertilizer, using values of number one (1 σ), two (2 σ) and three (3 σ) multiples of the standard deviation. The experiment was performed in rural area within the county of Sinop – MT, during the crop 2014/15. The experimental design used was based on the statistical quality control logic, to monitor the variables throughout the operational course. Has been collected 120 sampling points in total, 60 being collected per day (at random moments), for each seeding type in a period of two days, for each variant analyzed. The quality indicators were the seeder-fertilizer driving wheels skidding and overall field capacity, all variants being collected during the soybean seeding. The major probability of the occurrence of errors type I é presented to all the quality indicators which use value one (1 σ) and two (2 σ) as standard deviation multiple. The driving wheel skidding, both in the conventional seeding and in the direct seeding can be evaluated using the value multiple of the standard deviation number three (3 σ). The overall field capacity on the conventional seeding system can be evaluated using the value multiple of the standard deviation number three (3 σ). And, the direct seeding can be evaluated using the value multiple of the standard deviation number two (2 σ).