Study descriptives.

Note. y-YES, n-No; VAS = visual analogue scale; IB questionnaire = interpretation bias questionnaire, STAI-T/S = State Trait Anxiety Inventory – Trait/State; SST = Scrambled Sentence Test; PANAS = positive and negative affect scales; PAS = positive affect scale (from PANAS); NAS = negative affect scale (from PANAS); PANAS FS = fear subscale from PANAS; RT = reaction time; RIQ = Response to intrusions questionnaire, FNES = Fear of Negative Evaluation Questionnaire; EAST = extrinsic affective Simon task; NA = not available; * Murphy et al (2007) employed two independent benign groups that differed in that one group focused on emotional outcomes that were non-negative whereas the other group focused on pure positive outcomes. As this distinction was not made in the current meta-analysis both these training conditions were included under the ‘benign’ training condition.
1health status = healthy, anxious, depressed (note: this considers symptoms and clinical diagnoses); 2AS = ambiguous situations paradigm, WSAT = word sentence association task, ‘other’ paradigms included picture word interpretation, sentence completion (Lang et al, 2011), positive or negative valenced statements (Standage, 2010); 3training items/total items (ratio) = number of items in training direction divided by total number of items in task; 4word generation = were participants required to actively generate emotional meaning?; 5mode of presentation = were stimuli presented visually or auditory?; 6feedback administration = was feedback about response accuracy administered.