SciELO - Scientific Electronic Library Online

 
vol.66 issue1Effect of Addition of Hydrocolloids on Rheological Properties of the Brines Used to Prepare Cooked Ham author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

  • On index processCited by Google
  • Have no similar articlesSimilars in SciELO
  • On index processSimilars in Google

Share


Revista Facultad Nacional de Agronomía Medellín

Print version ISSN 0304-2847

Abstract

CORREA LONDONO, Guillermo. Analysis of Covariance as a Methodology to Control Confounding Variables. Rev. Fac. Nac. Agron. Medellín [online]. 2013, vol.66, n.1, pp.6981-6985. ISSN 0304-2847.

Some portion of the total variability in an experimental study can be explained by factors that are controlled and/or assigned by the researcher, and that are of his primary interest. Likewise, experiments usually involve factors that, despite their ancillary nature, also affect the response. Blocking is the most widely used mechanism to control the effect of ancillary factors. There are, however, situations in which the secondary source of variation is recognized only after the experiment has been started and/or in which its levels don't allow to group homogeneous experimental units. In such cases, it would be feasible to evaluate the use of analysis of covariance to achieve the same objectives that blocking does. In order to apply an adequate correction via analysis of covariance it is necessary to fulfill two conditions: viability and pertinence. Viability refers to the possibility to relate, by means of a regression model, a fraction of the variability of the response to the covariate. Pertinence has to do with the adequacy of the applied correction, taking into account that the elimination of the effect of the covariate doesn't extract some part of the treatment's effect. Viability is usually evaluated with the assistance of some statistical software. Pertinence, on the other hand, requires a conceptual approach.

Keywords : Internal validity; statiscal methods; general linear model; randomized complete block design.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )