SPSS Statistics

SPSS Statistics

Your hub for statistical analysis, data management, and data documentation. Connect, learn, and share with your peers! 

 View Only
  • 1.  Advice on coding dichotomous outcomes (0 vs 1) and deriving OR and RR in SPSS

    Posted Tue January 03, 2023 02:24 PM
    Hi, I'd be grateful for advice when deriving odds ratios & relative risk with 95% CI in SPSS. I'm looking at determinants of future hand pain and struggling with how the output is generated - hoping there is a workaround that doesn't involve recoding all my variables. 

    Outcome at Year 10: no pain (0), pain (1)
    Independent variable (for example): no pain at baseline (0), pain at baseline (1)

    When I derive the odds ratio and relative risk (prevalence of hand pain is >10%), I do this using the procedure Analyse>Descriptive statistics> Crosstabs. I place my independent variable in the row field and the outcome in the column field.

    The risk estimate table gives results for 'no hand pain at baseline' (coded 0) for the outcome 'no pain at Year 10' (coded 0). i.e. the OR (and RR) of having hand no hand pain at Year 10 if one has no hand pain at baseline. I'm guessing SPSS is set to look at 0 with 0. It would be much more informative to have the converse for my research question.

    Whilst I could manually work this out from the crosstabs table, I have several outcome and determinants (i.e scope for error and not time efficient!) and I don't know how I could (easily) work out the CI manually.

    I could get around this by recoding all my variables however I've already used some of these variables for other analyses and it would get VERY confusing when I return to the results in the future.

    Is there a way to get SPSS to look at variables coded 0 with 1, or even 1 with 1 (e.g. odds of HAVING hand pain at Year 10, should one HAVE hand pain at baseline)?

    Thank you!


    ------------------------------
    M Gul
    ------------------------------

    #SPSSStatistics


  • 2.  RE: Advice on coding dichotomous outcomes (0 vs 1) and deriving OR and RR in SPSS

    Posted Tue January 03, 2023 04:09 PM
    About all I can think of at the moment would be to recode into new variables to preserve the old ones.

    recode Pain (1=0) (0=1) into NoPain.
    value labels NoPain 0 "No" 1 "Yes".
    variable labels NoPain "Absence of Pain".
    etc...

    ------------------------------
    Rick Marcantonio
    Quality Assurance
    IBM
    ------------------------------



  • 3.  RE: Advice on coding dichotomous outcomes (0 vs 1) and deriving OR and RR in SPSS

    Posted Wed January 04, 2023 04:53 AM
    Thank you. I'll proceed with this.

    ------------------------------
    M Gul
    ------------------------------