Page 88 - Post-traumatic stress disorder in adults with mild intellectual disability: Screening, assessment, and brief intensive EMDR therapy

Chapter 4
Statistical analyses
All analyses were conducted in R (R Core Team, 2023, version 4.3.2) and
RStudio (Posit, 2024, version 2024.09.0). To determine the required sample
size, we calculated the number of permutations for this non-concurrent
multiple baseline between-subject design. For a robust randomisation test,
the number of possible starting points (k) must exceed 20, ensuring 1/k
(p < .05; Bulté & Onghena, 2009). With 11 participants and three starting
points, our study yielded 177 randomisation possibilities, resulting in 1/177
(p < .001), ensuring sufficient statistical power. To assess the test-retest
reliability within the baseline phase, the (Intraclass Correlation Coefficient;
ICC; Koo & Li, 2016) using a one-way random effects model to measure
consistency, was calculated for the BPI, TS-ID, and GAS.
To calculate the non-overlap of all pairs (NAP) for TS-ID, GAS, and BPI for
each participant, the SingleCaseES package (Pustejovsky et al., 2024) was
used. For the TS-ID and BPI, an improvement in symptoms was indicated
by a decrease in scores, and for the GAS, an improvement was indicated by
an increase. The baseline (phase A) was compared to the post-intervention
and follow-up phases (phase B). Missing values were excluded from the
analysis. The effect size was assessed based on the guidelines of Parker
and Vannest (2009).
Group-level randomisation tests were performed for the TS-ID, GAS, and BPI
(total and subscales) scores with the scan package (Wilbert & Luke, 2023)
to compare the baseline (phase A) with the post-intervention and follow-
up phases (phase B) using a t-statistic to measure phase differences.
Randomisation tests were conducted using the distributions of the data
based on random samples of n = 177 possible permutations. Missing
values were interpolated prior to analysis.
PTSD symptom frequency scores (TS-ID) were plotted for visual analysis
using the scplot package (Wilbert, 2023). PTSD diagnostic status (DITS-
ID) was visually analysed using a table. The use of involuntary care was
documented and described in the results section.
86