Basics of Minimum Information for A General Rule Regarding Human Beings:
1) Sample size: > 1000 ~3% error is close enough
2) Sample range: the general population, not a subgroup.
3) Sample source:
… a) reported (total bullshit) (surveys)
… b) observed (mostly bullshit) (observation testing – contrived)
… c) demonstrated (some bullshit) ( recorded testing – in the wild)
… d) discovered data (least possible bullshit) (finance/economics)
If you cannot make use of a sample size > 1000, from the general population, and using discovered data, then it’s almost guaranteed that you’re wrong – and wrong by a long shot.
The best data is an artifact of some other form of data collection.
Thus Endeth The Lesson
Source date (UTC): 2017-04-12 10:34:00 UTC
Leave a Reply