I am using the DHS Cambodia 2014 questions to assess the level of wealth within my sampled population. I am pretty new at using he DHS methodology and I apologise if the following questions sound a bit naive.

1. Continuous variables: I computed the number of HH members per sleeping room and then standardised this value, using the mean and standard deviation provided in the 2014 Cambodian DHS Excel file. Do you then add the results directly into the wealth score or should this continuous variable be categorised too? If categorised, how? Same questions apply to land area.

2. I see in the 2014 Cambodian DHS Excel file that the variables on the number of animals are to be transformed into new categorical variables: "No buffalo", "1-4 buffaloes", "5-9 buffaloes", ">9 buffaloes", etc. If a family has, say, 2 buffaloes, then does that mean that I need to sum the "has" coefficient of "1-4 buffaloes" with the "does not have" coefficients for "No buffalo", "5-9 buffaloes", ">9 buffaloes"?

3. Finally, my sample is exclusively located in rural areas. Do I need to compute the "common score" and then compute the "combined score" (using the linear regression criteria), or can I only use the rural score and directly apply the cut-off values to get my quintiles?

I was unable to find (or to understand) answers to the above questions in the provided documentation ("Steps to build the wealth index", etc.).

Thanks a lot.

Best regards,

Benjamin.]]>