The authors of this paper present a meta-analysis of the literature on expressive writing (EW) interventions for adolescent populations; they examine the effects of EW on particular subgroups of adolescents as well as the variability in effect size across studies as a function of specific characteristics of the participants and of the; they also investigate whether effect size is related to elements of the research design and assess the potential impact of biases or artifacts that are common in meta-analysis on effect size.
This meta-analysis evaluated the effects of the expressive writing intervention (EW; Pennebaker & Beall, 1986) among adolescents. Twenty-one independent studies that assessed the efficacy of expressive writing on youth samples aged 10–18 years were collected and analyzed. Results indicated an overall mean g-effect size that was positive in direction but relatively small (0.127), as well as significant g-effect sizes ranging from 0.107 to 0.246 for the outcome domains of Emotional Distress, Problem Behavior, Social Adjustment, and School Participation. Few significant effects were found within specific outcome domains for putative moderator variables that included characteristics of the participants, intervention instructions, or research design. Studies involving adolescents with high levels of emotional problems at baseline reported larger effects on school performance. Studies that implemented a higher dosage intervention (i.e., greater number and, to some extent, greater spacing of sessions) reported larger effects on somatic complaints. Overall, the findings suggest that expressive writing tends to produce small yet significant improvements on adolescents' well-being. The findings highlight the importance of modifying the traditional expressive writing protocol to enhance its efficacy and reduce potential detrimental effects. At this stage of research, the evidence on expressive writing as a viable intervention for adolescents is promising but not decisive. Publisher Abstract Provided