Summary

Let’s summarize what we’ve learned so far in this chapter.

• Three most important things: We discussed that the three most important things to consider are balance, randomization, and replication. A combination of balance and randomization is the best approach in many cases.

• Plan analyses before the experiment: We learned that everyone should plan their analyses before conducting an experiment. This is informative in many ways, but most of all, how we plan to analyze the data will influence the design of the experiment.

• Avoid dubious practices: We saw that once data is collected, we must avoid dubious practices such as p-hacking or HARKing. Instead, we should try to evaluate the preexisting hypotheses and be as neutral as possible when assessing if models are appropriate for explaining the response variables.

• Understand data: We discussed that data is just data. Regardless of the particular field, numbers and categories are just that: numbers and categories. The most important thing is that the data should be understandable.