Re: How to arrive at final wealth index score using PCA [message #30312 is a reply to message #30303] |
Thu, 31 October 2024 17:22 |
tahirandreas5
Messages: 1 Registered: October 2024
|
Member |
|
|
If you want to use the predict command after Principal Component Analysis (PCA), you will get factor estimates based on the specified component. These estimates can be used as an index of wealth.
However, it is important that the PCA model is set up correctly. For example, you need to choose the right number of components to keep. If you choose the wrong number of components, the results will be biased and the wealth index will be incorrect.
If factor scores are used without additional modifications such as standardisation or weighting, this will not be consistent with the methodology used by the household surveys (DHS) to produce the wealth index. The DHS often use more sophisticated methods that take into account different aspects of the data. For example, they use weighted components or additional statistical methods to make the index more accurate.
|
|
|