Keywords Running Head multiple comparison procedure Multiple Comparisons in Induction Algorithms David Jensen and Paul R. Cohen Experimental Knowledge Systems Laboratory Department of Computer Science Box 34610 LGRC University of Massachusetts Amherst, MA 01003-4610 413-545-3613 A single mechanism is responsible for three pathologies of induction algorithms: attribute selection errors, overfitting, and oversearching. In each pathology, induction algorithms compare multiple items based on scores from an evaluation function and select the item with the maximum score. We call this a ( ). We analyze the statistical properties of and show how failure to adjust for these properties leads to the pathologies. We also discuss approaches that can control pathological behavior, including Bonferroni adjustment, randomization testing, and cross-validation. Inductive learning, overfitting, oversearching, attribute selection, hypothesis testing, parameter estimation Multiple Com...

A graphical representation of the multivariate analysis of variance (MANOVA) is proposed. This representation, termed a "MANOVAMAP," shows the magnitude of MANOVA model effects and their statistical significance in an easily interpreted statistical graphic. The use and construction of MANOVAMAPs is discussed for an empirical example, and is compared with both a traditional MANOVA analysis of the data, and a traditional graphical analysis based upon centroid plots. Introduction The multivariate analysis of variance (MANOVA) model (see, for example, Bray & Maxwell 1985; Hand & Taylor 1987; Mardia, Kent & Bibby 1982) is a powerful and versatile marketing research technique. As a straightforward generalization of the analysis of variance (ANOVA), MANOVA allows the marketing researcher to test hypotheses involving differences in means of a set of p continuous dependent variables. Mean differences in these p variables can be tested across levels of S categorical independent variables (des...

day workshop on multilevel modelling will be

this document is as follows:

Abstract not found

• What data have we got? • What do we need to know about the repeated measures design?

Abstract not found

In classical analysis of variance, dispersion is measured by considering squared distances of sample elements from the sample mean. We consider a measure of dispersion for univariate or multivariate response based on all pairwise distances between-sample elements, and derive an analogous distance components (DISCO) decomposition for powers of distance in (0, 2]. The ANOVA F statistic is obtained when the index (exponent) is 2. For each index in (0, 2), this decomposition determines a nonparametric test for the multi-sample hypothesis of equal distributions that is statistically consistent against general alternatives.