Hedonic and eudaimonic pathways of self-forgiveness and suicide-related cognitions in trinidad and tobago: a quantile regression approach

Abstract Self-forgiveness is a multidimensional construct involving both hedonic processes, primarily related to fostering self-compassion and alleviating self-condemnation, and eudaimonic processes, focused on moral repair and value reorientation. While previous research has linked self-forgiveness to lower suicidality, little is known about whether these components operate similarly across different levels of suicide-related cognitions. This study examined the associations between the hedonic (esteem restoration; ERS) and eudaimonic (value reorientation; VRO) components of self-forgiveness and suicide-related cognitions in a general population sample from Trinidad and Tobago, using quantile regression. This cross-sectional study was conducted online via Qualtrics in 2023 with 869 English-speaking adults. Suicide-related cognitions were measured using the Brief Suicide Cognitions Scale, and self-forgiveness was assessed with the Self-Forgiveness Dual-Process Scale, comprising the VRO and ERS subscales. Quantile regression analyses were conducted at the 25th, 50th, and 75th percentiles of suicide-related cognition scores, controlling for demographic characteristics. Robust linear regression using Huber’s M-estimator was also performed to assess model stability. Quantile regression indicated that ERS was linked to significantly lower suicide-related cognitions at all examined quantiles, with the strongest associations observed at the 50th and 75th percentiles (β = –0.47, p < .001 for both). VRO (eudaimonic component) showed a significant negative association only at the 75th percentile (β = –0.13, p = .045). For comparison, robust linear regression using Huber’s M-estimator showed a significant negative association for ERS, whereas VRO was not statistically significant. This study underscores the importance of examining self-forgiveness beyond average effects, as its components may not operate uniformly across suicide-related-cognition levels. The use of quantile regression highlights these differences and points to the need for future longitudinal research to examine whether these associations reflect processes relevant to suicide prevention.