The study’s analysis incorporates advanced methods, such as full information maximum likelihood (FIML), to handle missing data. However, attrition rates were high, particularly for follow-up assessments (from 108 participants at baseline to 66 at T3). While FIML assumes data are missing at random (MAR), there is limited discussion about whether this assumption holds true for the sample. Participants who dropped out may systematically differ from those who completed the study, potentially biasing the results.
Could the authors provide additional evidence supporting the MAR assumption, such as comparisons of baseline characteristics between completers and non-completers? have sensitivity analyses been conducted to explore the robustness of findings under alternative missing data mechanisms?