SPSS Statistics

Hi everyone,

I'm working on an SPSS analysis and would really appreciate some guidance to ensure I’m using the correct statistical methods for my research questions.

Research Questions:

1. Are individuals with both ASPD and polysubstance use more likely to exhibit violent behavior than those with either risk factor alone?

2. What proportion of individuals diagnosed with ASPD report polysubstance use, and how does this compare to those without an ASPD diagnosis?

Variables I’m Working With:

• aspdph2: ASPD diagnosis (1 = yes, 0 = no)

• polysub_drugyear: Polysubstance use (1 = yes, 0 = no)

• VIOLENCE_BIN: Violent behavior (1 = yes, 0 = no)

• polysub_count: Number of substances used (continuous)

• Control variables: resage, ressex, ethnic4, nssec

• Weight variable: aspd_wt

• Sample size: ~6700

What I’ve Done So Far:

• Logistic regression with VIOLENCE_BIN as DV, and predictors:

  • aspdph2, polysub_drugyear, and their interaction (aspdph2 * polysub_drugyear)

• Crosstabs with Chi-square to compare rates of polysubstance use between ASPD vs non-ASPD groups

• GLM (Univariate) to compare polysub_count by ASPD status

• Applied weights via Data > Weight Cases... using aspd_wt

What I’d Like Help With:

• Am I using the right analyses to answer these questions?

• Is the interaction term in logistic regression appropriate for Q1?

• Is there a better way in SPSS to assess the group difference for Q2?

• Should I be considering mediation or other approaches?

• Any SPSS-specific guidance on running or interpreting these tests correctly?

Thanks in advance for any advice you can offer to this burnt out student xx