Longitudinal and experimental investigations of implicit happiness and explicit fear of happiness

ABSTRACT

Some individuals devalue positivity previously associated with negativity (Winer & Salem, 2016). Positive emotions (e.g. happiness) may be seen as threatening and result in active avoidance of future situations involving positivity. Although some self-report measures can capture emotions of happiness-averse individuals, they are not always capable of capturing automatic processing. Thus, we examined the association between implicitly-assessed happiness and explicit (i.e. self-reported) fear of happiness in three studies. In Study 1, participants completed the Fear of Happiness Scale (FHS) and an implicit measure of emotions at four-time points over approximately one year. The implicit measure required participants to choose which emotion (i.e. anger, fear, happiness, sadness, or none) best corresponded to 20 individual Chinese characters. In Studies 2 and 3, we utilized an experimental design, implementing a mood induction to emphasise the relationship between explicit fear of happiness and implicitly-assessed happiness. Participants completed the FHS and chose which emotion they believed the artist tried to convey in 20 abstract images. Results indicated that greater self-reported fear of happiness was related to reduced implicit happiness. Findings from these studies provide compound evidence that individuals who hold negative views of positivity may process implicit happiness in a devaluative manner.

KEYWORDS:

• Reward devaluation theory
• Happiness
• Implicit measure
• IMDES

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Although individuals are required to give a perception or attribution of what emotion is being conveyed in the ambiguous stimulus, we use the word “implicit” in the current study to describe the response that an individual gives in response to an ambiguous stimulus. This is in part due to the automatic, unconscious processing of these tasks such that participants often rely on their own emotions as a source of information.

2 Although the implicit measure in Study 1 used Chinese characters, we did not assess for whether participants were familiar with or spoke Chinese.

3 Given these data come from a convenience sample, coupled with a lack of a clear consensus on the parameters needed to estimate power for multilevel models (Lane & Hennes, 2017), no formal a priori power analysis was conducted.

4 The procedure of Study 2 was part of a larger study examining behavioural indices of avoidance. Thus, we did not include these face-valid questions to reduce potential demand effects on the behavioural tasks that followed the MIP and IMDES.

5 The scores for each of the 4 emotions (anger, fear, sadness, and happiness) can vary between 0 and 20. Thus, the neutrality would be indicated by a value of 0 for a given emotion score. The "none" response makes it possible for all emotion scores to be zero, as respondents can rate all pictures as expressing no emotions. Moreover, because respondents have to choose only one response option for each picture, including all 5 scores in the analyses would be redundant, as the "none" score can be determined once the other four scores are known.

6 FHS scores were not available for two participants due to them having missing data.

7 Given that the four emotions are not exclusive of one another, we also conducted a rmANOVA to investigate the effect of MIP on the endorsement of happiness. We included MIP as the between-subject factor and assessment period as the within-subject factors. Our findings support the results from our mixed ANOVA: there was neither a significant main effect of assessment period on implicit happiness, F(1, 226) = .39, p = .54, η_p²= .002, nor a significant interaction of assessment period and MIP on implicit happiness, F(1, 226) = .36, p = .55, η_p²= . 002. Including FHS as a covariate did not alter our findings.

8 We initially planned to run this study in person to better replicate Study 2; however, we were unable to do so due to COVID-19 interrupting in-person data collections. Thus, we ran the study online via Qualtrics.

9 Given that the four emotions are not exclusive of one another, we also conducted a rmANOVA to investigate the effect of MIP on the endorsement of happiness. We included MIP as the between-subject factor and assessment period as the within-subject factors. Our findings support the results from our mixed ANOVA: there was a main effect of assessment period on implicit happiness, F(1, 509) = 7.32, p = .007, η_p²= .01, and a significant interaction of assessment period and MIP on implicit happiness, F(1, 509) = 20.44, p < .001, η_p²= .04. Including FHS as a covariate revealed a significant interaction of assessment period on MIP, F(1, 507) = 5.02, p = .03, η_p²= .01, but all other main effects and interactions were not significant.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.