- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Posted 03-26-2016 08:36 AM
(3028 views)
How to detect influential observations in logistic regressiobn?
4 REPLIES 4
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
The documentation for the influence diagnostics says "the distributions of these diagnostic statistics are not known, so cutoff values cannot be given for determining when the values are large." To use these statistics, you plot them against an index and look for outliers.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Should i plot cook's D against predicted probabilities?
How to decide which one is outlier in absence of any cutoff value?
How to decide which one is outlier in absence of any cutoff value?
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
What is an outlier? The answer depends on you, the analyst. An influential observation can be an important observation to explain what's going on in your data or it might be an aberrant case where something went wrong. Influence diagnostics tell you which cases are worth investigating, nothing more.
If some cases are much more influential than the rest, then go to the source and seek answers about those.
PG
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Check MODEL statement option OBSTATS in PROC GENMOD .