“… - In studies that reported separate N2pc results for angry and happy facial expressions, effect sizes were pooled for each sample and averaged for inclusion in the meta‐analysis. However, in calculating the moderating effects of facial expression, anger‐induced and happy‐induced N2pc amplitudes were calculated separately for each sample.
- For one included mood state induction study (Liu et al., 2017), N2pc amplitude results were included only from neutral control conditions and were excluded from conditions featuring affectively valenced mood state inductions (e.g., after watching violent video).
- In studies that included distinct distressed groups (e.g., high trait socially anxious and chronic pain) (Holmes et al., 2013; Wang et al., 2020; Wieser et al., 2018) as well as healthy subgroups, N2pc amplitude effect sizes were calculated separately for each subgroup.
- In the moderator analysis for sample type, N2pc effect sizes were calculated separately for nondistressed/healthy versus distressed/less healthy subgroups (e.g., chronic pain and high social anxious).
- When results were reported for separate N2pc components (e.g., early vs. late N2pc (Feldmann‐Wüstefeld et al., 2011; Holmes et al., 2009; Holmes et al., 2013) or face image pair durations (e.g., 500 ms vs. 1,000 ms) from the same sample (Grimshaw et al., 2014; Osinsky et al., 2014; Yao et al., 2013), effect sizes were pooled and averaged to provide an overall effect size for the sample. However, in calculating moderating effects of time window, early (<250 ms) versus late (>250 ms) N2pc effects were calculated separately for each sample.
- In studies that featured interventions (Osinsky et al., 2014; Reutter et al., 2019), only pretreatment N2pc results were included for analysis to maintain consistency with designs of other included studies.
…”