Biased Adjustment for Bridging Studies Through Maximum Likelihood Technique
碩士 === 國立高雄大學 === 統計學研究所 === 96 === Efficiency is highly relative to the number of information, and therefore we may improve our efficiency by referring the published results. However, we should carefully use this kind of information, since publication bias. Publication bias arises because we often...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2008
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/u5ggmu |
| Summary: | 碩士 === 國立高雄大學 === 統計學研究所 === 96 === Efficiency is highly relative to the number of information, and
therefore we may improve our efficiency by referring the published
results. However, we should carefully use this kind of information,
since publication bias. Publication bias arises because we often
publish the significant results than non-significant conclusion, and
hence, the published data are truncated by alternative hypothesis.
It tends to give a tendency of over-estimate or under-estimate in a
long run. In this paper, we propose a method of using simple linear
regression model and maximum likelihood technique to adjust the
publication bias while merging the truncated data under bridging
study in order to make our estimation more efficient.
|
|---|