On the choice of computational unit in statistical analysis.

@article{Blomqvist1985OnTC,
  title={On the choice of computational unit in statistical analysis.},
  author={Nils Blomqvist},
  journal={Journal of clinical periodontology},
  year={1985},
  volume={12 10},
  pages={873-6}
}
It is stated that in trials where experimental units on different levels (sites, patients, etc.) are employed, the highest level unit should be used as computational unit when computing standard errors and in statistical inference. Using a lower level unit will underestimate the standard error and the level of significance (P-value). A numerical illustration is presented.