Testing non-additivity (interaction) in two-way ANOVA tables with no replication

Alin, Aylin; Kurt, S.
February 2006
Statistical Methods in Medical Research;Feb2006, Vol. 15 Issue 1, p63
Academic Journal
journal article
Testing for any significant interaction between two variables depends on the number of replicates in each cell of the two-way table and structure of the interaction. If there is interaction between two factors model of observations include interaction term and is called 'non-additive model' which makes interaction and non-additivity equivalent in terms of meaning. When there are several observations taken at each level combination of two variables, testing non-additivity can easily be done by usual two-way ANOVA method which cannot be used when there is only one observation per cell. For the cases with only one observation per cell, some methods have been developed starting with Tukey's one-degree-of-freedom test in which interaction is supposed to be the product of two factor's effects. There are other methods which are used for different structures of interaction when there is only one observation. In this paper, we review some of these tests. After presenting general methodology for the two-factor linear model with interaction effect and the general two-way ANOVA method when there are n > 1 observations per cell, we present some methods for testing non-additivity when there is only one observation per cell. Finally, we illustrate these methods on examples.


Related Articles

  • Distribution-free specification tests for dynamic linear models. DELGADO, MIGUEL A.; HIDALGO, JAVIER; VELASCO, CARLOS // Econometrics Journal;Mar2009 Supplement 1, Vol. 12, pS105 

    This article proposes goodness-of-fit tests for dynamic regression models, where regressors are allowed to be only weakly exogenous and arbitrarily correlated with past shocks. The null hypothesis is stated in terms of the lack of serial correlation of the errors of the model. The tests are...

  • THREE PARAMETRIC TECHNIQUES FOR CONTINGENCY TABLE ANALYSIS: A NONTECHNICAL COMMENTARY. Swafford, Michael // American Sociological Review;Aug80, Vol. 45 Issue 4, p664 

    This article presents a direct comparison of three "new" parametric techniques for contingency table analysis-linear probability, logit, and tog-linear modeling. The comparisons are rendered as nontechnical as possible, so that those familiar with ordinary regression will readily see how each of...

  • Mathematic coupling of data: a frequently misused concept. Squara, Pierre // Intensive Care Medicine;Oct2008, Vol. 34 Issue 10, p1916 

    The article focuses on the mathematical coupling of biological variables. It relates that all biological variables from a single patient are a priori linked to one another by overt or covert couplings resulting from anatomical, physiological, pathological, and/or therapeutic mechanisms. It...

  • Variable selection for the single‐index model. Efang Kong; Yingcun Xia // Biometrika;Mar2007, Vol. 94 Issue 1, p217 

    We consider variable selection in the single-index model. We prove that the popular leave-m-out crossvalidation method has different behaviour in the single-index model from that in linear regression models or nonparametric regression models. A new consistent variable selection method, called...

  • The Prep statistic as a measure of confidence in model fitting. F. Gregory Ashby; Jeffrey B. O'Brien // Psychonomic Bulletin & Review;Feb2008, Vol. 15 Issue 1, p16 

    In traditional statistical methodology (e.g., the ANOVA), confidence in the observed results is often assessed by computing the p value or the power of the test. In most cases, adding more participants to a study will improve these measures more than will increasing the amount of data collected...

  • PORTFOLIO THEORY: MEAN-VARIANCE. Board, John L. G.; Sutcliffe, Charles M. S.; Ziemba, William T. // Encyclopedia of Operations Research & Management Science;2001, p620 

    The article presents information on the mean-variance model of the portfolio theory. Portfolio problem involves the selection of an optimal set of investment assets by rational economic agents. The first formal specification of such a selection model was made by Markowitz. He defined a...

  • Optimal Fractional Dickey–Fuller tests. Lobato, Ignacio N.; Velasco, Carlos // Econometrics Journal;Dec2006, Vol. 9 Issue 3, p492 

    This article analyzes the fractional Dickey–Fuller (FDF) test for unit roots recently introduced by Dolado, Gonzalo and Mayoral (2002 Econometrica 70, 1963–2006) within a more general setup. These authors motivate their test with a particular analogy with the Dickey–Fuller...

  • A cautionary note on the analysis of randomized block designs with a few missing values. Mehrotra, Devan V. // Statistical Papers;Jan2004, Vol. 45 Issue 1, p51 

    The randomized block design is routinely employed in the social and biopharmaceutical sciences. With no missing values, analysis of variance (AOV) can be used to analyze such experiments. However, if some data are missing, the AOV formulae are no longer applicable, and iterative methods such as...

  • Main effects as a by-product in a fixed effect 2 × 2 design, analyzed with ANOVA. Landsheer, Johannes; Wittenboer, Godfried // Quality & Quantity;Oct2004, Vol. 38 Issue 5, p653 

    To study interaction effects, two sets of data are created for fixed effect ANOVA, both with combinatory effects of the two factors. In the first, both factors and their interaction contribute independently and directly to the dependent variable. In the second, each factor contributes indirectly...


Read the Article


Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics