Improving the accuracy and precision of disease assessments - selection of methods and use of computer-aided training-programs

The choice of assessment method has added significance considering the high cost of field research. The use of sophisticated statistical models and methods on assessment data with unknown variation has also come into question. In addition to using ANOVA, correlation analysis, and the coefficient of variation to evaluate and compare assessment methods and raters, least squares regression can be used to determine if there is a significant linear relationship between repealed assessments by the same rater (intra-rater repeatability) and whether there is a significant relationship between assessments by different raters (inter-rater reliability). Regression equation parameters (slope, intercept, standard error of the y-estimate, coefficient of determination) can be used to compare the accuracy and precision of methods and raters. The accuracy of raters can also be improved with proper training. Disease.Pro, a computerized disease assessment training program was developed to help raters improve their ability to estimate disease severity. After a one-hour training session with Disease.Pro, 70 out of 80 raters significantly improved their ability to estimate severity of late leaf spot of peanut.