Hi,

I was hoping someone might have a more elegant solution for the following issue. I'm running a ton of PCAs using a variety of model specifications. I'd like to hold all of the model specifications within a local, which I could use within a for loop. However, I'm running into a strange issue when creating this local. An example below:

Code:
*The following refer to variable names in the dataset:

local z_be bank_z childcarecenter_z conveniencestore_z conveniencestoreorsupermarket_z credit ///
                 union_z fastfood_z fedex_z firestations_z hospital_z landfill_z lawenforcement_z ///
                 library_z mobilehomeparks_z nursinghome_z pharmacy_z playground_z primarycareproviders_z ///
                 privateschool_z publicbuildings_z publichealthdepartments_z publicpark_z publicpool_z publicschool_z ///

local z_ses z_income z_mortgage z_rent z_unemployment z_poverty z_education

*I've tried the below with differing amounts of quotes and skew quotes;
*the following has gotten me closest to what I would expect.

local specifications "`z_ses'" "`z_be'" "`z_ses' `z_be'"

*Testing out if the above local holds my model specifications correctly:

local n 0
foreach x in `specifications'{
    di "Model `n': "
    di "`x'"

    local ++n
}
Oddly, when I do this, the above output spits out the individual elements of "z_ses" but correctly lists the rest of the model specifications (output below). I'm at a loss for why STATA is disaggregating the first term but not the second. How should I be creating this local to correctly hold different combinations of variables for model specifications?


Kind regards,
Max Griswold

Array