Browsing by Author "Robert L. Plaisted"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Design and Analysis of Onion Storage Trials
    (2007) Robert L. Plaisted; Horner, T.W.; Lana, E.P.
    This problem was designed to obtain information on methods of designing onion storage trials and the analysis of the data obtained from them. A storage trail was run two years, and during the last year at two locations, one designed for onion storage and the other the basement of the horticulture building, where the temperatures were high and the humidity low. The experiment was designed as a split plot with size of sample as the main plot and varieties as the subplot. Three sizes of sample, 25 bulbs, 50 bulbs, and 100 bulbs, and four varieties were used. Storage records were taken at monthly intervals during the last part of the storage period. All analyses were computed both on the basis of percentage weight loss and on the percentage transformed to angles by the arc sine transformation. Both sets of data were subjected to tests for non-homogeneity of variances and non-additivity, and to the test of the hypothesis that there was no binomial variation. A table is given which gives the estimated size of a real difference in weight loss expressed as a percentage of the mean that would be required to be detected at the five per cent probability level (Type I error) at three levels of probability of failure to detect any difference (Type II error) when the number of replications is four, six, or eight. The method is outlined for obtaining the same information for different numbers of replications.