Table 4.

Parameter estimates for the bivariate fixed RICLPMs modelling social media use and depressive symptoms (N = 2891).

Parameter	B	SE	95% CI	p	β
T1 correlation	0.03	0.02	−0.01 to 0.07	0.095	0.08
Correlated change	0.02	0.01	−0.01 to 0.04	0.233	0.04
Between-person correlation	0.02	0.02	−0.02 to 0.06	0.238	0.07
Cross-lagged paths
DepS → SoMe	0.13	0.06	0.01 to 0.25	0.039	0.08 to 0.10
DepS T3 → SoMe T5	0.15	0.19	−0.22 to 0.53	0.422	0.12
SoMe → DepS	0.01	0.03	−0.05 to 0.07	0.802	0.01 to 0.01
SoMe T3 → DepS T5	0.14	0.08	−0.03 to 0.31	0.097	0.18
Autoregressive paths
SoMe → SoMe	0.33	0.05	0.23 to 0.44	<0.001	0.34 to 0.37
SoMe T3 → SoMe T5	−0.08	0.14	−0.35 to 0.19	0.569	−0.08
DepS → DepS	0.12	0.10	−0.08 to 0.31	0.238	0.10 to 0.11
DepS T3 → DepS T5	0.16	0.13	−0.10 to 0.42	0.237	0.17
DepS T5 → DepS T6	0.56	0.07	0.42 to 0.70	<0.001	0.48

Note: B, unstandardized beta weights; β, standardized beta weights; CI, unstandardized confidence intervals; DepS, depressive symptoms; SoMe, social media use. Autoregressive and cross-lagged paths were set equal across time, excluding paths between T3 and T5.