This paper is based on my experience with advising students and colleagues doing research which involves econometrics. There are several common mistakes and misunderstandings about the nature of econometric work. This is because the issues to be discussed are either completely ignored, or else treated very superficially in conventional economic textbooks.
Common Mistakes:
Goal of research is to estimate and interpret a regression model.
Nonesense regression due to common trends/integrations.
Nonesense regression due to model selection.
Nonensense regression due to ignoring important determinants -- Length of Square on Area of square
Nonesense regression due to wrong functional form -- quadratic approximated by linear can yield highly misleading results
Nonensense regression due to ignoring causal chains. example of migration/yield
Outliers can lead to nonesense regression.
Observational Data versus Experimental Data and how it can lead to errors.
Quasi-Experimental Design -- care versus folly.
Exogeneity: Why Gini Coeff cannot be used in regression on Income Distribution.