Hi. I guess one thing you could do to figure out the average for a MULT RESPONSE set (dichotomous or not) would be to use the COUNT command on the variables that define the MR variable. It's just the total of the counted values divided by the number of cases. For example, here is syntax that creates some random data and two MULT RESPONSE sets.
dataset close all.
new file.
input program.
vector b(8).
vector c(5).
loop #i=1 to 25.
loop #j=1 to 8.
compute b(#j)=rnd(uniform(1)).
end loop.
loop #j=1 to 5.
compute c(#j)=1+rnd(uniform(3)).
end loop.
end case.
end loop.
end file.
end input program.
formats all (F1).
execute.
** The actual MULT RESPONSE variables, when tabulated, give you the total count for each across the dataset.
MULT RESPONSE GROUPS=
$b 'Binary variable set' (b1 b2 b3 b4 b5 b6 b7 b8 (1))
$c 'Categories 3 and 4' (c1 c2 c3 c4 c5 (3,4))
/FREQUENCIES=$b $c.
** The count per observation.
COUNT b=b1 to b8 (1).
COUNT c=c1 to c5 (3,4).
FRE VAR b c /FORMAT NOTABLE /STATISTICS MEAN.
That gives you the average count per observation (that is, row of the dataset). Multiply that by the number of cases and you reproduce the MR total count.
------------------------------
Rick Marcantonio
Quality Assurance
IBM
------------------------------