To ensure that your results are valid, consider the following guidelines when you collect data, perform the analysis, and interpret your results.
For more information on factors, go to Factors and factor levels, What are factors, crossed factors, and nested factors?, and What is the difference between fixed and random factors?.
Random samples are used to make generalizations, or inferences, about a population. If your data were not collected randomly, your results might not represent the population.
If the model does not fit the data, the results can be misleading. In the output, use the residual plots, the diagnostic statistics for unusual observations, and the model summary statistics to determine how well the model fits the data.