Effects of Group Forgiveness Intervention on Perceived Stress, State and Trait, Anger, Symptoms of Stress, Self-Reported Health and Forgiveness (Stanford Forgiveness Project)


Harris, A.H, Luskin, F.M.., Benisovich, S.V., Standard, S., Bruning, J., Evans, S. and Thoresen, C.  (2006)  Effects of a group forgiveness intervention on forgiveness, perceived stress and trait anger: A randomized trialJournal of Clinical Psychology.  62(6) 715-733.

August, 2001

To date there are a handful of successful forgiveness intervention studies.  This is the first study to attempt a forgiveness intervention with a large number of participants and that recruited a large cohort of men.  This randomized study of 259 adults compared effects of a 6 session (90 minutes each) manual-based cognitive behavioral intervention with assessment control group on perceived stress, state/trait anger, symptoms of stress, self reported health, forgiveness self-efficacy and forgiveness likelihood.  Participants completed baseline, posttest after the six week training, and 4 months follow-up assessments.  The intervention group received a forgiveness training developed by Dr. Luskin that drew upon the principles of cognitive disputation, mindfulness meditation and guided imagery.

The intervention focused on core components of forgiveness: taking less personal offense, blaming the offender less, and offering more personal and situational understanding of the offender and of oneself.  Participants (mean age = 41.8; 62% female; 31% minority) had at least one unresolved hurtful interpersonal experience that still elicited negative behavioral, cognitive and emotional responses.  Besides examining main treatment effects for each dependent variable, analysis also explored possible moderating variables.

For completers (N= 201), repeated measure ANOVAs revealed significant treatment effects for Perceived Stress with a post-test effect size (ES) = 0.71, women were consistently higher on stress.  Treatment but not controls experienced less stress at post and follow-up compared to baseline (p < 0.01).  Treatment vs. control effect size at follow-up was 0.54.  The treatment group also showed a 70% drop in the degree of hurt they felt from the baseline to the Follow-up six months later.

State and trait anger showed significant decreases for Treatment group participants (p = 0.011 & 0.010).  Positive results were seen in Trait Anger (long-term) with a post-test ES of 0.48 and State Anger (short-term) with ES of 0.40.  Marginally significant gender by condition effect (p = 0.087) suggested treated men improved more than women.  The treatment group increased their levels of optimism significantly more than the control group with a post-test ES of 0.27.

In addition two scales that looked at the physical symptoms of stress (Somatization and Other Scales of SCL90-R) showed significant decreases for treatment group participants.  For the 2 single-item measures of general health (MOS), no treatment effects were found.

For forgiveness self-efficacy, or how confident people felt about forgiving the offender, the results were significant (p = 0.000), and the effect size at follow-up was 0.60.  The intervention also increased participants willingness to choose forgiveness in other situations more than controls ( p = 0.000; effect size = 0.59).  In addition the intervention made participants more forgiving of their specific offender( p<0.03).

Implications are considerable, including the possibility that skills-based forgiveness training may prove effective in reducing anger as a coping style, reducing perceived stress and physical health symptoms, and thereby may help reduce allostatic load (e.g., immune and cardiovascular functioning) in daily living.