SUBJECTIVE COMPONENTS OF MILDEW ASSESSMENT ON SPRING BARLEY

The severity of mildew on barley is usually assessed visually and this leads to variation between different scorers. Field assessments by four assessors were analysed to determine the nature and degree of subjective discrepancies between assessors. Two inexperienced assessors failed to detect a major effect of nitrogen due to differences in the interpretation of a scoring system. A computer-based training programme was evaluated for standardising assessments, and was found to improve assessors' accuracy. Linear regression analysis was used here to resolve the error variance into components representing the accuracy and precision of the assessors. Plots of the cumulative differences between the estimate of disease severity by each assessor and the best estimate were used to display how the discrepancies varied with the level of disease. Some modifications to the barley field scoring system are suggested to improve comparability between assessors.