Skip to main content

Table 2 Parameter estimates for the multilevel models with sleep disturbances as the predictor variable

From: Sleep disturbances predict active suicidal ideation the next day: an ecological momentary assessment study

Models Est. SE t(df) p
Predictor
Model 1 (Active SI)
 Intercept 3.19 0.18 17.76 (72) <.001
 Sleep 0.18 0.09 2.04 (337) .042*
Deviance testa: χ2(2) = 4.63, p = .097 Quasi R2 (RI)b: 7.11% Quasi R2 (RS)c: 11.3%
Model 2 (Passive SI)
 Intercept 4.53 0.26 17.19 (72) <.001
 Sleep 0.10 0.11 0.95 (337) .343
Deviance testa: χ2(2) = 2.84, p = .240 Quasi R2 (RI)b: − 1.8% Quasi R2 (RS)c: 3.8%
Model 3 (Depression)
 Intercept 5.94 0.21 28.26 (72) <.001
 Sleep 0.15 0.13 1.16 (337) .245
Deviance testa: χ2(2) = 0.46, p = > .500 Quasi R2 (RI)b: −1.1%Quasi R2 (RS)c: − 0.8%
Model 4 (MIS Sum Score)
 Intercept 8.16 0.24 33.54 (72) <.001
 Sleep 0.46 0.27 1.67 (321) .096
Deviance testa: χ2(2) = 0.33, p = > .500 Quasi R2 (RI)b: 1.9% Quasi R2 (RS)c: − 1.4%
Model 4a (MIS Item 1)
 Intercept 2.16 0.10 22.52 (72) <.001
 Sleep 0.33 0.12 2.67 (321) .008**
Deviance testa: χ2(2) = 0.61, p = > .500 Quasi R2 (RI)b: −1.2% Quasi R2 (RS)c: 1.6%
Model 4b (MIS Item 2)
 Intercept 1.47 0.06 23.91 (72) <.001
 Sleep −0.03 0.11 -0.26 (321) .799
Deviance testa: χ2(2) = 0.90, p = > .500 Quasi R2 (RI)b: 1.4% Quasi R2 (RS)c: 2.1%
Model 4c (MIS Item 3)
 Intercept 2.86 0.10 29.52 (72) <.001
 Sleep 0.15 0.12 1.25 (321) .212
Deviance testa: χ2(2) = 0.39, p = > .500 Quasi R2 (RI)b: −1.7% Quasi R2 (RS)c: −1.4%
Model 4d (MIS Item 4)
 Intercept 1.66 0.08 22.17 (72) <.001
 Sleep 0.00 0.11 0.03 (321) .974
Deviance testa: χ2(2) = 0.65, p = > .500 Quasi R2 (RI)b: 2.7% Quasi R2 (RS)c: 3.1%
  1. Notes. All reported models are random intercept (RI) models. Predictor variable is sleep disturbances. n (level 1 – daily means) = 443; n (level 2 - participants) = 74; Est. = unstandardized regression coefficient; SE = standard error; Model 1 = active suicidal ideation as outcome variable; Model 2 = passive suicidal ideation as outcome variable; Model 3 = depression as outcome variable; Model 4 = impulsivity with MIS sum score as outcome variable; Model 4a = MIS Item 1 as outcome variable; Model 4b = MIS Item 2 as outcome variable; Model 4c = MIS Item 3 as outcome variable; Model 4d = MIS Item 4 as outcome variable;*p < .05, **p < .01; aDeviance tests: random intercepts vs. random slopes model, bQuasi R indicates the percentage of residual variance in the outcome variable that is explained when adding the models’ level 1 predictors compared to the baseline models (RI models), cQuasi R indicates the percentage of residual variance in the outcome variable that is explained when adding the models’ level 1 predictor compared to the baseline models (non significant RS models)