Addendum to: Translational Psychiatry (2017) 7, e1001; doi:10.1038/tp.2016.263; published online 17 January 2017

After publication, the authors determined that regression tests are the appropriate post hoc tests for clarifying the direction of the three MANOVA results that found significant associations between functional connectivity changes and long-term clinical outcomes. In each of the three regression tests, greater increases in functional connectivity were associated with improved long-term clinical outcomes, as ascertained by positive beta weights for the following regressors: longitudinal positive psychotic symptoms (β=0.157, t(14)=2.79; P=0.015), longitudinal affective symptoms (β=0.187, t(14)=2.36; P=0.035) and subjective long-term recovery (β=0.13, t(14)=2.56; P=0.024). These tests replace the correlation tests originally reported in the article.