Lain Iwakura
Active Member
- Aug 26, 2021
- 213
- 225
- 31
- Country
- United States
- Faith
- Eastern Orthodox
- Marital Status
- In Relationship
Also, for outliers, if it's a data point in a data set you can get away with saying that it's some fluke that can't be explored analytically. If it's any multiple of data points, especially clustered, a specific and defined reason should be included as to why the outliers should be excluded.
Treating a large and valuable dataset as an outlier is bad analytics practice - as not only do you need a specific and well-defined reason for doing so - but your in the realm of meta-analytics between data sets and if you cannot establish specific methodological failures or phenomenal differences, then you have no reason to dismiss the data set at all.
Let's not forget it was Israel's datasets, being one of the most comprehensive ones, that was among those consulted for major issues, including the Myocarditis issue - although it was first discovered in VAERS, the comprehensiveness of Israel's datasets were used to figure out a concrete statistical connection and occurrence rate.
Treating a large and valuable dataset as an outlier is bad analytics practice - as not only do you need a specific and well-defined reason for doing so - but your in the realm of meta-analytics between data sets and if you cannot establish specific methodological failures or phenomenal differences, then you have no reason to dismiss the data set at all.
Let's not forget it was Israel's datasets, being one of the most comprehensive ones, that was among those consulted for major issues, including the Myocarditis issue - although it was first discovered in VAERS, the comprehensiveness of Israel's datasets were used to figure out a concrete statistical connection and occurrence rate.
Upvote
0