Animal behavior research, like other sciences, adheres to a rigorous methodology of design and execution. It begins with careful delineation of the research questions, objectives and hypotheses. Dependent and independent variables are identified, including behaviors to be measured or manipulated. A research protocol is structured, which casts the variables and animal subjects into the proper experimental design, prescribes appropriate scales of measurement and designates valid parametric or non-parametric statistical analyses. Data collection begins with the choice of appropriate sampling methods and equipment to insure validity, accuracy and reliability. The most robust and powerful data analyses, commensurate with the parameters of the data set, are conducted and interpreted relative to the research hypotheses and questions.