FDD – A USEFUL MS EXCEL ADD-IN FOR FITTING DISCRETE DISTRIBUTIONS TO DISEASE INCIDENCE, WEED, INSECT AND NEMATODE FREQUENCY DATA

Document Type : Research Article

Authors

Abstract

An add-in for MS Excel 2007 was developed to fit discrete distributions to the frequency of disease incidence, weed, insect and nematode in the form of count per sampling unit data. Probability distributions such as binomial, Poisson, negative binomial and beta binomial are discrete distributions, which are respectively appropriate for describing uniform, random and aggregated or clustered binary data such as disease incidence or insect/nematode/weed count per quadrate observations. The fit discrete distribution (FDD) program is a very user-friendly and flexible add-in and can take data in the form of either Table of raw observations or a frequency Table. The program estimates distribution parameters and their standard errors using a maximum likelihood procedure and determine the expected (theoretical) values of the distribution for any distribution from its list. It also calculates both chi-square and log-likelihood ratio goodness of fit statistics to test the null hypothesis of goodness of fit and plot the expected frequencies according to all used distributions against observed values in one single bar graph for comparison purposes. Application of the program in studying spatial pattern of plant pests is discussed.

Keywords