0

I had a question on my questionnaire that asked living arrangements and if participants selected option B or C they got asked another question. All data on SPSS for those who selected A is coming up as a '.' as they obviously did not get to see the next question.

Should I just write all missing answers as a number e.g. 3 and then label the number 3 as 'ignore' or 'NA' in variable view or is there something else I can do?

Please note I'm a 3rd year university student who is awful at SPSS so I don't know all the technical terms! Hope someone can help.

Thanks.

Alice Willoughby
  • 1
  • 1
    It depends what you want to do with the data, but basically you don't necessarily have to do anything - the '.' in numeric variables means a missing value, and will not be used in analyses. – eli-k Mar 29 '20 at 08:09

1 Answers1

0

As mentioned in the comments, it really depends on how you intend on using your data. That said, one approach is to assign values to your system missings to increase flexibility. I'll riff an example based on your description.

Q1) Please describe your living arrangements. a. I live alone. b. I live with family/roommates and am the head of household. c. I live with family/roommates, but am not the head of household.

[If Q1=a Then skip] Q2) How many other members, excluding yourself, do you live with? a. 1 b. 2 c. 3 d. 4 or more

Q3) Are any of these [Q2 count] members non-family roommates? a. Yes b. No

Since Q2 is a count of people in household (excluding self), you might recode those living on their own to actually answer the question.

RECODE Q2 (SYSMIS=0).
EXE .

Or if there was a need to differentiate (either because not all SYSMIS should be 0 or because you want to keep track of why they are 0). In this case you can toggle those missing values on/off as needed.

IF (Q1='a') Q2=0.001 .
EXE .
ADD VALUE LABELS Q2 .001 'Lives Alone' .
MISSING VALUES Q2 (.001) .

For Q3, it really is more of an N/A. In that case you might choose an arbitrary value (-1, 99, etc) for tracking purposes and always keep those missing values set.

If (Q1='a' AND Q3=$SYSMIS) Q3=99 .
EXE .
ADD VALUE LABELS Q3 99 'N/A, lives alone' .
MISSING VALUES Q3 (99) .
user45392
  • 610
  • 7
  • 16