ERIC Number: ED173209
Record Type: RIE
Publication Date: 1977
Reference Count: 0
Final Report for Dynamic Models for Causal Analysis of Panel Data. Approaches to the Censoring Problem in Analysis of Event Histories. Part III, Chapter 2.
Tuma, Nancy Brandon; Hannan, Michael T.
The document, part of a series of chapters described in SO 011 759, considers the problem of censoring in the analysis of event-histories (data on dated events, including dates of change from one qualitative state to another). Censoring refers to the lack of information on events that occur before or after the period for which data are available. Unless censorship is dealt with, researchers are likely to make erroneous inferences about the change process. The report considers several approaches to estimation when event-histories are censored. A constant rate (Poisson) model is considered because the methodological issues are more easily understood. Models in which the rate of an event depends on exogenous variables or time and in which there are multiple kinds of events are also analyzed. The report then discusses approaches to estimation based on maximum likelihood (ML), pseudo-maximum likelihood, the method of moments, and recent work by statisticians on methods that make weak parametric assumptions. The conclusion is that an important advantage of the ML approach to the censoring problem is that it is easily extended to different data structures and different models. (Author/KC)
Publication Type: Reports - Research
Education Level: N/A
Sponsor: National Inst. of Education (DHEW), Washington, DC.
Authoring Institution: Center for Advanced Study in the Behavioral Sciences, Stanford, CA.; Stanford Univ., CA.
Note: For related documents, see SO 011 759-772