Authenticity and self-esteem across temporal horizons

Abstract

Extending research on optimal self-esteem and authenticity, three studies tested the hypothesis that authenticity would be a stronger predictor of self-esteem levels when time was perceived as limited as opposed to open ended. Study 1 provided a cross-sectional examination of the relationship between authenticity, future time perspective, and self-esteem in an adult sample, and Studies 2 and 3 assessed this relationship using repeated measures methodologies across both the short term and long term in college student samples. Results supported the hypothesis that authenticity would be a stronger predictor of self-esteem levels when time was perceived as limited. Across studies, individuals who felt inauthentic reported lower levels of self-esteem when they perceived time as limited.

Keywords:

• authenticity
• future time perspective
• self-esteem
• personality

Notes

1. Additional analyses on a subsample (n = 256) of participants from Study 1 are reported by Davis and Hicks (2013).

2. Simple slope analyses of the age × FTP interaction were conducted via recentering FTP at ±1 SD from the mean. Age was a stronger predictor of self-esteem when participants perceived time to be limited (b = 0.033 and p < 0.001) compared to when participants perceived time to be open ended (b = 0.013 and p = 0.003).

3. When the 13 multivariate outliers were included in the regression analysis, the pattern and significance of the results were unchanged. The main effects of authenticity (b = 1.018 and p < 0.001), FTP (b = 0.469 and p < 0.001), and age (b = 0.027 and p < 0.001) were very consistent with the analysis excluding the outliers and continued to significantly predict self-esteem. Importantly, the authenticity × FTP interaction (b = −0.199 and p = 0.002) remained significant. The age × FTP interaction also remained significant when the outliers were included in the analysis (b = −0.005 and p = 0.020).

4. When the same analysis is run without PA and NA as covariates, the interaction effect is somewhat attenuated (b = −0.15 and p = 0.054).

5. When the same analysis is run without PA and NA as covariates, the significance of the interaction effect is somewhat attenuated, though the size of the beta is essentially unchanged (b = −0.29 and p = 0.077).