Skip to main content

Table 4 Model comparisons between the null and the alternative models for the study variables (N = 16)

From: Reducing suicidal ideation among Turkish migrants in the Netherlands and in the UK: the feasibility of a randomised controlled trial of a guided online intervention

Variable Model BF10 Ratio
BSS (T0 vs T3) Null model 1.00 a
  Time 50.41 1.00
  Group 0.46 0.01
  Time + group 27.27 0.54
  Time + group + time × group 11.89 0.23
BSS (T1 vs T2) Null model 1.00 a
  Time 1.32 1.00b
  Group 0.60 0.45
  Time + group 0.81 0.61
  Time + group + time × group 0.46 0.35
BDI (T0 vs T3) Null model 1.00 a
  Time 99.84 1.00b
  Group 1.06 0.01
  Time + group 127.09 1.27c
  Time + group + time × group 70.34 0.70
BHI (T0 vs T3) Null model 1.00 a
  Time 14.44 1.00b
  Group 0.54 0.04
  Time + group 8.30 0.57
  Time + group + time × group 7.27 0.50
PSWQ (T0 vs T3) Null model 1.00 a
  Time 2.85 1.00b
  Group 0.57 0.20
  Time + group 1.72 0.60
  Time + group + time × group 1.29 0.45
EQ5SD (T0 vs T3) Null model 1.00 a
  Time 0.43 1.00b
  Group 0.77 1.79c
  Time + group 0.33 0.76
  Time + group + time × group 0.14 0.32
  1. T0, baseline (pre-test); T1, bi-weekly measures of BSS at week 2; T2, bi-weekly measures of BSS at week 4; T3, post-test; BF10, Bayesian factor grading the intensity of the evidence supporting the alternative model against the null model. Ratio: This column represents the ratio (the likelihood) of the effect of time against the group, time and group and the interaction models. The time model is the denominator. The BF10 of each model has been divided by the BF10 of the time model in order to calculate the ratio of each model when compared with the time model
  2. aThe ratio for the null model was irrelevant
  3. bThe ratio of time against time is always 1
  4. cWhen the ratio is between 0 and 1, there is a weak evidence supporting the alternative model against the time model. When it is greater than 1, that means there is a stronger evidence supporting the alternative model against the time model