Between 1986 and 1990, a series of thinnings were made in previously unthinned first-rotation stands on former arable land located in the southern half of Sweden. The aim was to evaluate the effects of season and urea treatment on the frequency of infection of stumps of Norway spruce (Picea abies (L.) Karst.) by the root-rot fungus Heterobasidion annosum (Fr.) Bref. Untreated stumps, resulting from 60 thinnings (22-100 stumps each, altogether ca 3000 stumps) made at different times of year, were investigated 3-24 months after cutting to determine whether they were infected with H. annosum. On average only 2% of the stumps from thinnings made in November-February were infected, whereas the incidence of infection among stumps thinned in June-July was 34%. Two methods of treating stumps with urea to prevent stump infection by H. annosum after thinning were evaluated in terms of effectiveness. The freshly cut stumps were treated with a 20% urea solution, transformed to a gel by adding 0.2% carboxymethyl cellulose, or with a 30% urea solution. On average, the reduction in infection rate obtained was 62% with the first method and 85% with the latter. In a separate study involving a concentration series of urea, there was a considerable drop in protection efficiency, from 89% to 58%, when the concentration was decreased from 30% to 15%.
Statistical analysis of experiments giving dichotomous data and using a design with repeated observation of the same Ss or with matched groups is discussed. A minor extension of Cochran's Q test is presented and the importance of exactly specifying the null hypothesis when using this test is emphasized. The possibility of using the F test for dichotomous data is investigated in computer-simulated experiments. The results indicate that for the examples chosen, the F test as a rule gives at least as good results as the Q test with regard to the Type I error. The approximation to the theoretical significance level is good except in some cases where there are very strong deviations from F test assumptions.
Computer simulation experiments are performed to investigate the applicability of different tests of treatment and interaction effects i n mixed two-way designs with dichotomous data. The main results are that the F test can be used with good approximation as a test of equal treatment effects, except for strongly asymmetric populations, and that the F test may be used with some caution as a test of no interaction.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.