Usually I use Karup King formula to do it, but I observed that this method leads to some irregularities at the higher ages. My idea was to combine two methods: Lagrange method for the age range [1- x] and the Karup King method for the age range [ x - 80 or higher]. the age x (connection point) is defined in the order to insure a better junction between the two obtained curves. I need some orientation and references?
I found a very interesting section of this book*, starting on page 694 (Interpolation of grouped data). They provide various approaches and compare results.
* The methods and materials of demography: Volume 2 Henry S. Shryock Jacob S. Siegel Elizabeth A. Larmon. 1980
Thank you for your answear. I have this book. it's very interesting. it explains the different methods that we can use for data interpolation. what i'm looking for is which method use for interpolate data adapted to the adridged life table structure: 0, 1-4, 5-10, 10-15 ....etc. The quality of the interpolated data is generally less regular for the lower or the higher extreme.
This is not my area of study, but there are some potential solutions from other disciplines. The following references discuss approaches based on densities (binning) for determining cutoff levels. This is keeping with the spirit of your approach to partitioning age categories, while ensuring sufficient samples in each bin. I hope these may be helpful to you in thinking about your problem.
Scott, David W. 1992. Multivariate density estimation. Theory, practice and visualization. New York: John Wiley & Sons, Inc.
Freedman, David, and Persi Diaconis. 1981. On the histogram as a density estimator: L2 theory. Probability Theory and Related Fields 57:453–76.
Shimazaki, Hideaki, and Shigeru Shinomoto. 2007. A method for selecting the bin size of a time histogram. Neural Computation 19:1503–27.
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology
- Age Groups