which events contribute the most risk for same-day SI?

Two models, one with all negative events, one with non-significant negagtive events trimmed

Results

    Suic_Good   Suic_Good
    B CI p   B CI p
Fixed Parts
(Intercept)   5.06 4.63 – 5.48 <.001   5.04 4.62 – 5.46 <.001
A lot of work at home, job, or school.   0.06 -0.16 – 0.29 .571    
A lot of demands made by others   -0.01 -0.30 – 0.27 .930    
Someone you care about got sick or was injured   -0.36 -1.05 – 0.33 .309    
Argument with significant other   0.94 0.64 – 1.24 <.001   0.94 0.64 – 1.24 <.001
Argument with boss/authority figure   0.60 -0.17 – 1.37 .133    
Argument with someone else   0.16 -0.20 – 0.51 .389    
Was ignored by someone.   0.39 0.15 – 0.64 .002   0.41 0.17 – 0.65 .001
Was criticized by someone.   0.25 -0.03 – 0.53 .087   0.30 0.04 – 0.56 .028
Tension or problem with your child(ren).   1.05 0.29 – 1.81 .008   1.01 0.26 – 1.77 .011
Was late to an appointment.   -0.36 -0.92 – 0.19 .203    
Someone cancelled plans.   0.02 -0.37 – 0.40 .932    
Someone broke a promise to you.   0.45 0.09 – 0.81 .017   0.42 0.07 – 0.77 .021
Unexpected expense (e.g., repairs, ticket) or other money problem.   -0.15 -0.51 – 0.20 .405    
Problem with transportation or technology.   -0.23 -0.59 – 0.12 .202    
Failed to understand something.   0.37 0.05 – 0.69 .027   0.35 0.03 – 0.67 .034
Unable to complete plans, work, etc.   0.22 -0.02 – 0.45 .071   0.19 -0.04 – 0.42 .104
Had a minor accident (e.g., broke something).   0.11 -0.45 – 0.66 .711    
Random Parts
σ2   1.665   1.662
τ00, subject   2.104   2.117
Nsubject   53   53
ICCsubject   0.558   0.560
Observations   986   986
R2 / Ω02   .623 / .622   .619 / .618

Model comparision

non-sig chi-square means model with fewer factors is better

Df AIC BIC logLik deviance Chisq Chi Df Pr(>Chisq)
EVENTS_trimmed 10.00 3464.84 3513.78 -1722.42 3444.84
EVENTS 20.00 3475.83 3573.71 -1717.92 3435.83 9.01 10.00 0.53

Visualizing regression weights

sjp.lmer(EVENTS_trimmed,type="fe")
## Computing p-values via Kenward-Roger approximation. Use `p.kr = FALSE` if computation takes too long.

which events contribute the most risk for next-day SI?

Two models – one controlling for same-day SI, one not

Results

    Suic_Good_NX   Suic_Good_NX
    B CI p   B CI p
Fixed Parts
(Intercept)   5.38 4.93 – 5.83 <.001   3.62 3.17 – 4.07 <.001
Argument with significant other   0.36 0.03 – 0.69 .036   0.01 -0.31 – 0.33 .961
Was ignored by someone.   0.14 -0.12 – 0.39 .293   0.04 -0.21 – 0.29 .770
Was criticized by someone.   -0.10 -0.38 – 0.17 .467   -0.19 -0.47 – 0.08 .166
Tension or problem with your child(ren).   -0.21 -1.01 – 0.59 .606   -0.52 -1.29 – 0.25 .187
Someone broke a promise to you.   0.04 -0.33 – 0.42 .817   -0.09 -0.46 – 0.27 .609
Failed to understand something.   0.10 -0.23 – 0.43 .564   -0.05 -0.37 – 0.28 .779
Same-day SI       0.34 0.28 – 0.41 <.001
Random Parts
σ2   1.805   1.708
τ00, subject   2.317   0.980
Nsubject   50   49
ICCsubject   0.562   0.364
Observations   940   928
R2 / Ω02   .591 / .590   .610 / .609

Model comparision

Less straightforward here

df AIC
EVENTS_NX 9.00 3391.43
EVENTS_NX_C 10.00 3270.04