Functions for analysis of rate changes in sequential events. Appropriate data are the times of observation of well defined events, such as equipment failures or deaths.
|Packaged:||2013-05-28 11:12:16 UTC; root|
|Built:||R 3.0.0; ; 2013-05-28 11:12:19 UTC; unix|
1 2 3 4
The basic model is of events that are independent and uniformly distributed in time. This assumption underlies the use of the exponential distribution in the generalized linear models produced by EIglm. In general, it is expected that deviations from uniformly distributed intervals follow a function that is expressed in one or more of the predictors. A predictor that is included by default is the times of the occurrence of the events. This will reveal linear changes in the rate. Other predictors may test for non-linear or cyclic changes in rate.
Event interval analysis can be a useful alternative to Poisson based analysis when few events occur in the counting intervals. Its advantage in power to detect changes in rate diminishes as the number of events increases.
This type of analysis was known as event history analysis, but the phrase is now usually applied to the analysis of the antecedents of events, whether sequential or not.
Maguire, B.A., Pearson, E.S. & Wynn, A.H.A. (1952) The time intervals between industrial accidents. Biometrika, 39(1/2): 168-180.
Whitworth, W.A. (1901) Choice and chance. (3rd ed.) Cambridge: Deighton Bell.
Lemon, J. (2014) The analysis of event rates using intervals. The Quantitative Methods for Psychology, 10 (1), 68-76.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.