analysing categorical data with very high of outcome of interest

[Nazarius Mbona Tumwesigye]

Hello to all.
we are looking at possibilities for further analysis of knowledge of Family planning (FP) methods which stands at >97% in a certain population. my quick thoughts are that probably no need for further analysis as the knowledge is universal. we just keep it descriptive. Otherwise we end up with perfect prediction for many zeros in certain categories. What are your thoughts on this? Any technique for further analysis you can think of?

[Felix Bittmann]

This paper here gives an overview. Hope this is what you mean: https://www3.nd.edu/~rwilliam/stats3/RareEvents.pdf

[Fei Wang]

In addition to Felix's advice about rare-event analysis, you may also explore other measures of FP knowledge, for example a measure for the amount of knowledge (a continuous variable), for further analysis.

[Nazarius Mbona Tumwesigye]

This paper here gives an overview. Hope this is what you mean: https://www3.nd.edu/~rwilliam/stats3/RareEvents.pdf

thanks a lot

[Nazarius Mbona Tumwesigye]

In addition to Felix's advice about rare-event analysis, you may also explore other measures of FP knowledge, for example a measure for the amount of knowledge (a continuous variable), for further analysis.

thanks a lot