Better exploration of data to identify common statistical problems

Sometimes some scientific work does not meet the assumptions required by the statistical techniques they have used (i.e. something known as violation of assumptions). Some of the violations to the assumptions have low impact on the results or the conclusions, however others increase the errors of type I and II, which potentially affects the interpretations and the conclusions. Most of these violations can be resolved if better exploration are applied to the data. In this presentation, some tips are presented to make a better exploration of the data through the use of some examples obtained from studies carried out with the Neotropic birds. In particular, I will focus on the discussion of tools that serve to detect extreme data, heterogeneity of variances, collinearity, dependence on observations, problems with interactions, and the correct type of relationship between response and explanatory variables. Better exploration of data would improve the quality of ecological research and any applied recommendations that it produces.