abstract: Social media has been extensively researched, and its effects on well-being are well established. What is less studied, however, is how social media affects romantic relationships specifically. The few studies that have researched this have found mixed results. Some researchers have found social media to have a positive influence on relationship outcomes, while other have found social media to have a negative influence. In an attempt to reconcile these discrepancies, the current thesis study explored possible mediators between social media use and relationship health outcomes which, to my knowledge, has not been investigated in previous literature. Three moderators were explored: type of social media use (active use versus passive use), relationship-contingent self-esteem, and social comparison orientation. The baseline portion of the study had 547 individuals, recruited from Arizona State University’s SONA system as well as Amazon’s Mechanical Turk, who were in a romantic relationship for at least three months; the follow-up portion of the study had 181 participants. Results suggest that women who passively use social media exhibit a negative association between hours per day of social media use and baseline relationship satisfaction. Men who passively use social media exhibited a negative association between hours per day of social media use and follow-up relationship satisfaction, as well as a negative association with baseline commitment. While relationship-contingent self-esteem did not moderate the association between hours per day of social media use and relationship health, it was positively related to both men and women’s baseline relationship satisfaction and baseline commitment. Social comparison orientation (SCO) produced minimal results; women low on SCO exhibited a negative association between social media use and baseline relationship satisfaction, and higher SCO for men was associated with lower baseline commitment. Finally, exploratory post-hoc mediation models revealed that relationship comparisons mediated the association between hours per day of social media use and baseline relationship, as well as baseline commitment, for both men and women. Previous research supports the findings regarding passive social media use, while the findings regarding relationship-contingent self-esteem and relationship comparisons add new findings to the romantic relationship literature. / Dissertation/Thesis / Masters Thesis Psychology 2019

Social ångest på sociala medier är ett relativt outforskat område där fortsatt forskning underlättas av psykometriskt tillfredsställande självskattningsskalor. Följande studie syftade till att översätta Social Anxiety Scale for Social Media Users (SAS-SMU) och psykometriskt utvärdera den avseende dess latenta dimensioner, interna konsistens samt konvergerande respektive divergerande validitet i ett svenskt urval. Därutöver undersöktes samband mellan total användning och passiv respektive aktiv användning av sociala medier och social ångest, eftersom tidigare studier visat en tendens till passiv användning. Tre faktorer erhölls för SAS-SMU och resultaten visade hög intern konsistens samt att skalan konvergerade med SPIN, och divergerade med OCI-R, SWSL, PHQ-9 och GAD-7. Skalans tredje faktor erhöll dock svagare samband med SPIN jämfört med de andra faktorerna. Social ångest associerades med passiv användning i högre utsträckning än aktiv, samt med längre genomsnittlig användning av sociala medier. Resultaten skiljer sig från tidigare studier där signifikanta samband endast erhållits mellan passiv användning och social ångest. / Measurement scales are needed to facilitate further research on social anxiety in the context of social media. This study aimed to translate the Social Anxiety Scale for Social Media Users (SAS-SMU) and evaluate its psychometric properties in a Swedish sample regarding its underlying dimensions, internal consistency, converging and diverging validity. Furthermore, relationships between passive, active and total use of social media and social anxiety were explored, since previous studies have shown a tendency towards passive use. Three factors were retained for SAS-SMU with satisfactory internal consistency. SAS-SMU converged with SPIN and diverged with OCI-R, SWLS, PHQ-9 and GAD-7. However, the association between its third factor and SPIN was weaker compared to the other factors. Results indicated that higher levels of social anxiety were more strongly associated with passive use than active use, as well as longer general social media use. This is at odds with previous studies only reporting significant correlations between social anxiety and passive use of social media.