Authentic self-expression on social media is associated with greater subjective well-being

Social media users face a tension between presenting themselves in an idealized or authentic way. Here, we explore how prioritizing one over the other impacts users’ well-being. We estimate the degree of self-idealized vs. authentic self-expression as the proximity between a user’s self-reported personality and the automated personality judgements made on the basis Facebook Likes and status updates. Analyzing data of 10,560 Facebook users, we find that individuals who are more authentic in their self-expression also report greater Life Satisfaction. This effect appears consistent across different personality profiles, countering the proposition that individuals with socially desirable personalities benefit from authentic self-expression more than others. We extend this finding in a pre-registered, longitudinal experiment, demonstrating the causal relationship between authentic posting and positive affect and mood on a within-person level. Our findings suggest that the extent to which social media use is related to well-being depends on how individuals use it.


Supplementary
Model 2, we added personality extremeness as well as the main effects of self-reported personality to control for the fact that personality traits are known to be associated with Life Satisfaction1-4.  Table 4).
The robustness of our findings was tested by calculating various types of distance and similarity measures. We consider the robustness of Quantified Authenticity by calculating Manhattan distance, cosine similarity and correlational similarity as below.
The second distance measure we utilized is a Manhattan Distance measure. Manhattan Distance uses a grid-like measure of distance between two points. Distance is calculated as the sum of the horizontal and vertical distances between two points over the dimensions of the vectors,  This conceptualization assumes that individuals have an inherent drive to self-enhance on social media, and therefore any deviation from the true self is a reflection of self-idealizing behavior.
However, given the existing literature on socially desirable personality profileswhich show that, on average, high Openness, Extraversion, Conscientiousness and Agreeableness, but low Neuroticism are desirable.8 As a robustness check, here we isolate the extent to which deviation from self-reported personality presents as idealization by separating idealization from selfdepreciation on Life Satisfaction, and illustrating their effects on Life Satisfaction separately.
Specifically, we first calculated the difference between the self-reported and predicted personality traits such that positive and higher values indicate higher levels of self-idealization, and negative and lower values indicate higher levels of self-deprecation (i.e. subtracting the selfreported from the predicted score for Openness, Conscientiousness, Extraversion and Agreeableness, and subtracting the predicted score from the self-reported score for Neuroticism).
A score of zero would indicate that there is no deviation in the self-reported and predicted personality-profiles of a participant.

As Supplementary Figures 4A (Likes-based model) and 4B (Language-based model)
show, the effect of authentic self-expression on Life Satisfaction was particularly pronounced for the normative self-enhancing aspect of Quantified Authenticity. That is, higher levels of selfenhancement towards socially desirable personality traits were associated with lower levels of Life Satisfaction.

D. Supplementary Notes -Study 2
Positive and Negative Affect Additional Analyses. The pre-registration plan for Study 2 is available on OSF. 9 We note an error in the pre-registration plan which resulted from an oversight of one of the authors when setting up the online survey: the pre-registration incorrectly referred to wrong citation as the positive and negative affect measure. As mentioned in the manuscript, participants completed the 16-item Brief Mood Introspection Survey instead of the PANAS scale which was pre-registered. Past research has found that these two scales are highly correlated10, predict similar outcomes11-13, and can be combined with acceptable internal reliability14. Additionally, we returned to the PANAS-X handbook to review the similarities between the two scales. We found that there were seven items (four items for positive affect: lively, happy, content, active; three items for negative affect: sad, jittery, nervous) which were shared between the two scales. We extracted these items to create second measures positive affect and negative affect and test the replication of our effects on only these items in the Results section below.
Additionally, we considered how the 7-item PANAS measure performed against our hypotheses. At t0, the measure of positive affect as measured by the BMIS had good consistency (alpha = .81), as did the measure of negative affect (alpha = .82). The 4-item measure of PANAS positive affect also had internal reliability (alpha = .77). The 3-item measure of negative affect based on PANAS had lower internal reliability than the BMIS measure, but was still at an acceptable level (alpha = .64). Additionally, we found that the subset of scales was positively related to the overall positive affect scale reported in the manuscript (r = .93, p < .001) and the subset of negative affect was strongly and positively related to the overall negative affect scale reported in the manuscript (r = .86, p < .001).
Using the PANAS subset, we replicated our main experimental outcomes such that within participants differences were found. In the authentic week, participants reported significantly higher positive affect compared to their self-reported affect in the self-idealized Results of key outcome variables on condition assignment for week 1 are presented in Supplementary Table 8. Results of key outcome variables on condition assignment for week 2 are presented in Supplementary Table 9 with robustness checks for demographics (age and gender) and main effects of the Big Five personality traits.