# DataONE:Stats/RecurrentEventRegression

### From OpenWetWare

(Difference between revisions)

(initial content) |
Current revision (18:16, 4 August 2010) (view source)m (initial content) |
||

Line 1: | Line 1: | ||

- | Normal linear regression not good enough. Was exploring censored negative binomial, but how about Andersen-Gill survival analysis instead? | + | Normal linear regression not good enough for modelling citations, especially when the papers are published across a wide span of years. Was exploring censored negative binomial, but how about Andersen-Gill survival analysis instead? |

Some papers in Mendeley: | Some papers in Mendeley: | ||

Line 8: | Line 8: | ||

To be explored later, when I need it again... | To be explored later, when I need it again... | ||

+ | |||

+ | [[Category:DataONE]] |

## Current revision

Normal linear regression not good enough for modelling citations, especially when the papers are published across a wide span of years. Was exploring censored negative binomial, but how about Andersen-Gill survival analysis instead?

Some papers in Mendeley:

Mendeley citation generator

To be explored later, when I need it again...