Counting across several survey questions - spss

Question

I am working with survey data and have been stuck on an issue for several hours that i am not sure how to resolve.
Respondents in my survey, were asked how frequently they had been involved in negative situations at their work place. Now, this was asked in three different questions as
(1) How frequently have you had negative experiences with your boss?
(2) ... with colleagues
(3) ... clients
Possible answers for each questions were one of the following (1) daily, (2) several times a week, (3) once a week, (4) several times each month, (5) once a month, (6) less than once a month, (7) never and (8) would prefer not to answer.
What I would like to do:
I want to count how many of my respondents that have answered between (1) - (5), (6) - (7) and (8) as missing.
Importantly, I would like to know how many unique people that have been involved in negative situations. That is, as long as you have selected between (1) - (5) for any of the answers, I would like to count that as one (If you have selected between (1) - (5) for all three or two of the questions, it should still be counted as one). Also, how many that have selected between (6) - (7) must be counted in the same manner. 8 should be counted as missing.
Do any of you know how I can resolve this? Been stuck on this for a long time now!
Hoping for answers.

Answer 1

You can use the syntax below. I have named your three variables Boss, Colleague and Client. Copy the syntax into a syntax file in SPSS, change the variable names appropriately and run the syntax.
**This syntax will creates five new variables from initiall Boss Colleague and Client variables.
*The first three variables will recode the three variables based on if they are 1-5=1; 6-7=2 and 8=0.
*The last two variables Harrassed_15 checks if each case has 1-5 in at least one of the first three variables.
The variables Harrassed_67 checkes if each case has at least 6-7.
RECODE Boss Colleague Client (1 thru 5 = 1) (6 thru 7 = 2) (8 = 0)
INTO Boss_Grouped Colleague_Grouped Client_Grouped.
VALUE LABELS Boss_Grouped Colleague_Grouped Client_Grouped 1 '1-5' 2 '6-7' 8 'Missing'.
FREQUENCIES Boss_Grouped Colleague_Grouped Client_Grouped.
TEMPORARY.
COUNT Harrassed_15=Boss_Grouped Colleague_Grouped Client_Grouped(1).
COUNT Harrassed_67=Boss_Grouped Colleague_Grouped Client_Grouped(2).
FREQUENCIES Harrassed_15 Harrassed_67.
COMPUTE Harrassed_15 = (((Boss_Grouped=1) + (Colleague_Grouped=1) + (Client_Grouped=1))>0).
COMPUTE Harrassed_67 = (((Boss_Grouped=2) + (Colleague_Grouped=2) + (Client_Grouped=2))>0).
FREQUENCIES Harrassed_15 Harrassed_67.
*Alternatively.
*This is alternative to above. Creates only two variables without showing how many belonged to each group.
*It produces same results.
*If you don't want to produce additional variables in your dataset then remove the TEMP.
TEMPORARY.
COMPUTE Harrassed_15 = (((Boss<=5) + (Colleague<=5) + (Client<=5))>0).
COMPUTE Harrassed_67 = ((((Boss=6) OR (Boss=7)) + ((Colleague=6) OR (Colleague=7)) + ((Client=6) OR (Client=7)))>0).
FREQUENCIES Harrassed_15 Harrassed_67.