correlations by year group and post for 2 variables

[HEB1]

Dear users,

My data is attached.

I want to run correlations for each year, for all 3 groups between GDP and zscore.

for example: for year 2002   I want to calculate correlation for ordinal adoption=1 between GDP and zscore.

for year 2002   I want to calculate correlation for ordinal adoption=2 between GDP and zscore.

for year 2002   I want to calculate correlation for ordinal adoption=3 between GDP and zscore.

for year 2003   I want to calculate correlation for ordinal adoption=1 between GDP and zscore.

for year 2003   I want to calculate correlation for ordinal adoption=2 between GDP and zscore.

for year 2003   I want to calculate correlation for ordinal adoption=3 between GDP and zscore.

do that for all years in data and get the results in a nice table. 

Another question: If I want to add conditions for groups with ordinal_adoption1 or 2 so the correlation will only consider post=1 in the years - how to add that variable (in group ordinal_adoption=1 post=0 always) ? for example:

for year 2002   I want to calculate correlation for ordinal adoption=2 between GDP and zscore, only when post=1.

for year 2002   I want to calculate correlation for ordinal adoption=3 between GDP and zscore.only when post=1.

Thanks

[PaigeMiller · Accepted solution]

I want to run correlations for each year, for all 3 groups between GDP and zscore.

Use PROC CORR with a BY statement.

proc corr data=have;
    by year ordinal_adoption;
    var gdp zscore;
run;

Another question: If I want to add conditions for groups with ordinal_adoption1 or 2 so the correlation will only consider post=1 in the years - how to add that variable

Repeat the above code adding a WHERE statement.