Many phytopathological studies and routine activities in crop improvement depend on assessments of disease severity.  Severity has been linked to non-biased indicators, such as disease incidence (Seem, 1984), or quantified with sophisticated equipment, such as image analysis (Brodny et al., 1986; Lindow and Web, 1984) and infrared radiation (Nutter et al., 1993; Pedersen and Fiechtner, 1980). These procedures, however, are normally too expensive or too labor-intensive for large-scale disease evaluations.  For this reason, severity of foliar diseases is generally estimated visually.

To improve visual estimation of disease severity, scales have been developed for numerous diseases (see Campbell and Madden, 1990; Hau and Kranz, 1990; Kranz, 1988 for discussions).  Horsfall & Barratt (1945), through application of the Weber-Fechner law to visual assessment of disease severity, proposed that the human eye would estimate high and low disease severities with greater precision than mid range severities.  To correct this problem, they suggested that scale increments should be logarithmic rather than linear.

 Several studies have attempted to test the application of the Weber-Fechner law to the assessment of disease severity by examining inter-assessor variability at different severity levels (see Campbell & Madden, 1991; or Kranz, 1988).  Literature to date supports the hypothesis that precision of estimation is greatest at low and high extremes of disease severity, but the relationship between severity and precision is also affected by other factors.  Hebert (1982) argued that the relationship between visual acuity and true severity is not necessarily logarithmic.  It has also been shown that precision and accuracy are affected by the shape of the plant part being assessed (Forbes and Jeger, 1987) and the number of the lesions (Sherwood et al., 1983).  Systematic errors in assessment have been called 'illusions' and discussed in more detail by Campbell & Madden (1991).

Based on the premise that the Weber-Fechner law is generally applicable to visual disease estimation, or at least that there is a systematic relationship between severity and variance a logarithmic transformation of data, or use of a logarithmic  scale should correct problems of uneven variances.  This conclusion, however, has seldom been tested in the field.  Hebert (1982) pointed out that there is little experimental evidence that demonstrates that a Horsfall-Barratt scale is more effective in the field than no scale at all.  O’Brian & van Bruggen (1992) compared a Horsfall-Barratt scale with qualitative scales, but not with simple estimation of percentage infection.  Forbes and Korva (1994) compared a Horsfall-Barratt scale and direct percentage on potato plants in the field infected with  Phytophthora infestans.  They found that evaluators tended to linearize the scale and therefore direct percentage estimation was more accurate. 

Brodny, U., R. R. Nelson, L. V. Gregory (1986): The residual and interactive expressions of "defeated" wheat stem rust resistance genes. Phytopathology 76, 546-549.

Campbell, C. L., L. V. Madden. (1990): Introduction to Plant Disease Epidemiology. John Wiley & Sons, New York City.

Forbes, G. A., M. J. Jeger (1987): Factors affecting the estimation of disease intensity in simulated plant structures. Journal of Plant Diseases and Protection 94, 113-112O.

Forbes, G. A., J. T. Korva (1994): The effect of using a Horsfall-Barratt scale on precision and accuracy of visual estimation of potato late blight severity in the field. Pla. Path. 43, 675-682.

Hau, B., J. Kranz. (1990): Mathematics and statistics for analysis in epidemiology, p. 12-52. In: J. Kranz (ed.). Epidemics of Plant Diseases. Springer-Verlag, Berlin.

Hebert, T. T. (1982): The rationale for the Horsfall-Barratt plant disease assessment scale. Phytopathology 72.

Kranz, J. (1988): Measuring plant disease, p. 35-50. In: J. Kranz and J. Rotem (eds.). Experimental techniquesin Plant Disease Epidemiology. Springer-Verlag, Berlin.

Lindow, S. E., R. r. Web (1984): Quantification of foliar plant disease symptoms by microcomputer-digitized video image analysis. Phytopathology 73, 520-524.

Nutter, F. W. J., M. L. Gleason, J. H. Jenco, N. C. Christians (1993): Assessing the accuracy, intra-rater repeatability and inter-rater reliability of disease assessment systems. Phytopathology 83, 806-812.

O'Brien, R. D., A. H. C. Van Bruggen (1992): Accuracy, precision, and correlation to yield loss of disease severity scales for corky root of lettuce. Phytopathology 82, 91-96.

Pedersen, V. D., G. Fiechtner. (1980): A low-cost compact data acquisition system for recording visible and infrared reflection from barley crop canopies, p. 71-75. In: P.S. Teng and S.V. Krupa (eds.). Crop Loss Assessment Misc. Publ. 7 Agricultural Experiment Station. University of Minnesota, St. Paul, MN.

Seem, R. C. (1984): Disease incidence and severity relationships. Annu. Rev. Phytopath. 22, 137-150.

Sherwood, R. T., C. C. Berg, M. R. Hoover, K. E. Zeiders (1983): Illusions in visual assessment of Stagonospora leaf spot of orchardgrass. Phytopathology 73, 173-177.