Bonjour/Hi SAS fellows, it's me again !
Based on the interpretation of the Standard Error (SE) on the SAS proc stdrate Output .... (see solution/post : PROC STDRATE and Standard Error (Output)); is it appropriate to use this estimate (i.e. SE) for the calculation of reduced/centered (directly) standardized rates (proportions) ?
More precisely, one STD rate shall correspond to the entire population of interest (i. a. the USA or Canada) and the others to STD rates of subgroups of the same population (such as states or provinces of the same population). These standardized rates of the subgroups (stratum) will be the ones that we seek to reduce and center using the SE computed for the entire standardized population computed by the SAS proc stdrate. These calculations would clearly correspond to an additional step for the purpose of a visual presentation by means of an Excel sheet and based on the well known formula of a centered/reduced Z value.
Does it make sense ? Thanks in advance for your input.
