Skip to main content
. Author manuscript; available in PMC: 2018 Nov 1.
Published in final edited form as: Prev Sci. 2017 Nov;18(8):964–975. doi: 10.1007/s11121-017-0822-0

Table 2.

Intervention Effectiveness on Teachers and Students

Mean Score from T1 Raw Data Mean Score from T2 Raw Data Model-Based Inference

Intervention Control Intervention Control Difference [95%CI] p
Teacher Outcomes Mean (SD) Mean (SD) Mean (SD) Mean (SD) (SE)

EBP Knowledge 35.50 (13.39) 39.39 (12.98) 46.86 (17.28) 33.21 (16.11) 13.68 (5.80) [2.29, 25.08] .019
EBP Use-observed 2.18 (.35) 2.22 (.23) 2.16 (.23) 2.03 (.22) .16 (.07) [.02, .30] .024
EBP Use-self-reported 2.70 (.67) 2.47 (.58) 2.97 (.67) 2.72 (.53) .02 (0.19) [−.36, .40] .919

Child Outcomes
Social Competence 2.34 (.63) 2.35 (.67) 3.04 (.72) 2.53 (1.09) .66 (.22) [.23, 1.08] .003
Emotion Regulation 2.24 (.73) 2.14 (.73) 2.96 (.87) 2.29 (1.21) .79 (.25) [.30, 1.28] .002
Prosocial Skills 2.44 (.73) 2.55 (.74) 3.12 (.70) 2.78 (1.08) .54 (.21) [.12, .96] .012
Behavior Problems
Externalizing 1.30 (1.51) .96 (1.14) 1.02 (1.50) 1.31 (1.57) −.50 (.30) [−1.10,.11] .11
Internalizing .87 (1.32) 1.04 (1.20) .59 (1.16) .55 (1.23) .11 (.32) [−.53, .75] .74

Note. There were no significant differences at baseline. The mean scores in table are average scores from raw (non-imputed) data. Baseline Means and SDs for the full sample were 37.26 (13.26) for EBP knowledge, 2.20 (0.29) for EBP Use-observed, 2.59 (0.64) for EBP Use-self reported, 2.34 (0.64) for total child social competence, 2.20 (0.73) for emotion regulation, 2.49 (0.73) for prosocial skills subscales, 1.15 (1.37) for child externalizing problems, and 0.95 (1.27) for child internalizing problems. The model-based inferences are based on linear mixed effect models and the ITT principle from imputed data, with the exception for the EBP Use-observed. Observations were limited to 63 teachers/classrooms (excluded 6 school leaders/administrators). Also, 10 teachers (5 from intervention and 5 from control schools) who were not in schools during the observation periods at T1 and T2 were excluded because we could not impute observational data for cases without any observation. For all study outcomes, using imputed and non-imputed data yield similar results.