Table 2 DSEM results of the short-term effects of social media use with friends (SMU) on friendship closeness (FCL) based on the ESM assesments.

From: Some socially poor but also some socially rich adolescents feel closer to their friends after using social media

 

Model 1.1 (baseline model)

Model 1.2 (H1 & H2)

b

β

p

95% CI

b

β

p

95% CI

Fixed effects

Within-Person

FCLt−1→ FCLt(AR_FCL)

0.263

.263

.000

[.249, .276]

0.263

.263

.000

[.250, .277]

SMUt* → FCL t (ST_SMI_FCL)

−0.160

−.041

.000

[−.053, −.029]

−0.160

−.041

.000

[−.053, −.029]

Between-Person

SMU & FCL

0.043

.125

.010

[.024, .227]

0.036

.110

.018

[.007, .213]

LON & FS

    

−0.113

−.232

.000

[−.324, −.134]

FS → ST_SMI_FCL (H1a vs. H1b)

    

−0.059

−.063

.166

[−.189, .061]

LON → ST_SMI_FCL (H2a vs. H2b)

    

0.109

.139

.022

[.004, .264]

FS → FCL

    

0.363

.176

.000

[.075, .271]

LON → FCL

    

−0.292

−.169

.001

[−.268, −.068]

FS → SMU

    

0.037

.091

.045

[−.013, .195]

LON → SMU

    

0.002

.006

.459

[−.100, .109]

 

Model 1.1 (baseline model)

Model 1.2 (H1 & H2)

σ2

p

95% CI

σ2

p

95% CI

Variance (Random Effects)

Within-Person

FCL

1.977

.000

[1.947, 2.008]

1.977

.000

[1.947, 2.008]

R2 FCL

0.127

.000

[0.119, 0.135]

0.127

.000

[0.120, 0.135]

Between-Person

FCL

1.730

.000

[1.498, 2.015]

1.609

.000

[1.399, 1.869]

ST_SMI_FCL

0.354

.000

[0.277, 0.448]

0.346

.000

[0.274, 0.437]

AR_FCL

0.034

.000

[0.028, 0.041]

0.033

.000

[0.027, 0.041]

R2 ST_SMI_FCL

   

0.032

.000

[0.004, 0.086]

R2 FCL

   

0.076

.000

[0.031, 0.135]

R2 SMU

   

0.011

.000

[0.001, 0.040]

  1. Social media use (SMU) was dummy coded (i.e., 0 = no social media use with friends, 1 = social media use with friends); ST_SMI_FCL = short-term social media-induced change in friendship closeness; AR_FCL = the autoregressive effect of friendship closeness; FS = Friendship Support; LON = Loneliness; bs are unstandardized; βs are standardized using the STDYX Standardization in Mplus55; p-values are one-tailed Bayesian p-values27; significant fixed effects are depicted in bold.
Back to article page