A replication and conceptual evaluation of commonly used positive psychology interventions

What makes people happy is an increasingly important question for clinical practice and public health. The major line of research and intervention in this area, positive psychology, focuses on furthering knowledge about the factors which improve individuals’ levels of happiness and nurture the growth of character strengths. However, despite the growth in positive psychology research over the past decade, replication studies are lacking and the cross-cultural applicability of interventions has not been thoroughly investigated. Three studies are presented in this thesis, all of which relate to the landmark research conducted by Seligman, Steen, Park and Peterson (2005) on positive psychology exercises (PPEs). The purpose of Study 1 was to evaluate the efficacy of three PPEs and a control exercise in an Australian rather than American population. Consistent with the original study, an internet based randomised trial with four groups was used, in which each exercise was completed over a one-week period, with follow up measurements taken up to six months after completing the initial exercise. The findings of the original study were not fully replicated. Specifically, although all groups showed an increase in happiness levels and a decrease in depression levels over time, there was no differential effect between the PPEs and the control exercise. The aim of Study 2 was to examine whether the results from Study 1 might be attributable to problems in measuring happiness. In Study 1 the Authentic Happiness Inventory (AHI) was used to capture changes in happiness levels as it was designed by Seligman et al. to be sensitive to upward changes in Seligman’s (2002) three domains of happiness: pleasure, engagement and meaning. However, supporting literature regarding the psychometric properties of the AHI is lacking. In Study 2 discriminant content validity techniques were used to investigate how well the AHI represents the aforementioned psychological constructs. Study 2 showed that expert judges could not unambiguously allocate the AHI items to the intended constructs, indicating that the AHI demonstrates poor discriminant content validity. The purpose of Study 3 was to investigate the efficacy of the PPEs in an n-of-1 design using a more widely validated measure of subjective well-being, the Positive and Negative Affect Schedule (PANAS). The use of n-of-1 designs has been recommended when the emphasis is on examining within-person changes rather than between-group differences. Over the 9–10 week intervention period, no significant changes in happiness or depression were produced and apart from a small positive interaction between one of the PPEs and positive affect, there was no differential effect between PPEs and the control exercise. These largely non-significant findings raise doubts about the clinical appropriateness of these PPE interventions. Overall the results of this thesis demonstrated poor support for Seligman et al.’s (2005) study findings, raising questions not only about the interventions and measures promoted by Seligman et al., but also about the underlying theoretical concepts. Although the usefulness of PPEs in clinical settings appears limited, it is possible that there is scope for their use in a public health context. However before a public health use is pursued further, investigation is required into what the ‘active’ elements of PPE interventions are and whether the effects of these elements might be attributed to more general psychological theories of behaviour change. On this basis, it is recommended that future research efforts focus on addressing the cross-cultural and public health relevance of positive psychology interventions.