. 2016 Jul 20;16:254. doi: 10.1186/s12888-016-0969-4

Table 3.

Post-hoc multiple comparisons on time effect for ITT and Protocol Completer sample

	Intent to treat (n = 11)						NET-Protocol Completers (n = 10)
	Pre vs. Post	p	Pre vs. FU	p	Post vs. FU	p	Pre vs. Post	p	Pre vs. FU	p	Post vs. FU	p
PTSD	−2.97_18.63	.008	−4.32_19.48	<.001	−1.54_18.85	.140	−3.15_17.99	.006	−4.48_18.85	<.001	−1.55_18.85	.137
Borderline Symptoms	−2.27_18.92	.035	−2.66_19.45	.015	-.58_18.93	.584	−2.30_17.89	.034	−2.67_18.47	.015	-.55_18.47	.592
Depressive Symptoms	−3.70_19.22	.001	−3.98_19.83	.001	-.55_19.28	.558	−3.60_18.03	.002	−3.88_18.68	.001	-.54_18.68	.596
Dissociative Symptoms	−2.27_18.65	.035	−2.43_18.94	.025	-.33_18.57	.744	−2.28_17.89	.035	−2.44_18.22	.025	-.34_18.22	.739
Quality of Life	3.07_19.42	.006	2.99_19.98	.007	.16_19.46	.877	3.05_18.29	.007	2.98_18.89	.008	.152_18.89	.881

Note. Mixed effects models were used to disclose time effects. Within these models, contrasts were used to analyze post-hoc multiple comparisons. Displayed are t-values and degrees of freedom subscripted