10. Statistical Conclusion Validity
Most data are generated in part by chance processes. It is impossible to know in advance exactly what the data will show, and whether, if the data were generated a second time, they will differ at least a bit in ways that can not be anticipated. This uncertainty in the data means that any summary statistics computed and any conclusions that follow will be subject to uncertainty as well.