Skip to main content

On the failure to replicate past findings regarding positive affirmations and self-esteem (Pages 49-61)

Journal of Contextual Behavioral Science (JCBS)

Volume 16, April 2020, Pages 49-61


Maureen K. Flynn, Michael J. Bordieri



Previous research has established a differential effect of positive affirmations, whereby they may be helpful for individuals with high trait self-esteem and possibly iatrogenic for individuals with low self-esteem (Wood, Perunovic, & Lee, 2009). The current studies are replications of Wood et al. (2009) study. The aim of Study 1 was to replicate this study by examining the efficacy of a positive affirmation intervention on mood and state self-esteem and to see if trait self-esteem made the intervention more or less efficacious. The aim of Study 2 was to attempt to conceptually replicate and extend the original study by examining the efficacy of positive affirmation and values writing interventions on mood, self-esteem, and goal completion. We also examined whether trait self-esteem and psychological inflexibility moderated the relationship between condition and outcome variables.


For Study 1, participants were randomly assigned to either the positive affirmation or no statement condition. Outcome measures were completed immediately following the intervention. For Study 2, participants randomly assigned to the positive affirmations or values writing intervention. Outcome measures were completed immediately following the intervention and at a three-day follow-up.


Participants completed this analogue study in a university setting.


Approximately 225 individuals participated in Study 1 and 237 in Study 2.

Main outcome measures

The main outcome measures of Study 1 were state self-esteem and mood measures. The outcome measures for Study 2 were state self-esteem, mood, goal competition, and challenge of self-reported goal.


Results from Study 1 revealed a failure to replicate as there were no difference between conditions on all outcome variables and trait-self esteem did not interact with condition to predict any of the outcome variables. Findings from Study 2, again, yielded a failure to replicate, as there were no differences between conditions on mood, state self-esteem, self-reported level of goal challenge, and goal completion.


State self-esteem did not moderate the effect of a positive affirmation intervention, failing to replicate the findings of Wood et al. (2009). Implications regarding the failure to replicate the original study are discussed.

This article is restricted to ACBS members. Please join or login with your ACBS account.