Skip to main content
. 2019 Oct 26;91(4):e853–e865. doi: 10.1111/cdev.13334

Table 3.

Preliminary Results, Standardized (n = 1,629)

SMU intensity SMU problems Attention deficits Impulsivity Hyperactivity
β SE p β SE p β SE p β SE p β SE p
Wave 2a .39 .06 < .001 .84 .05 < .001 .27 .06 < .001 −.05 .07 .476 .09 .06 .117
Wave 3a .84 .06 < .001 .35 .06 < .001 .67 .07 < .001 .20 .06 .001 .09 .07 .163
Girlsb .42 .09 < .001 .22 .08 .007 −.07 .09 .445 −.18 .09 .046 −.02 .09 .842
Prevocational educationc .26 .09 .003 .45 .09 < .001 .25 .10 .010 .28 .09 .003 .26 .10 .007
Native ethnic backgroundd .25 .12 .034 −.03 .012 .831 .19 .13 .150 .02 .14 .871 .26 .11 .021
ICCe .803 .899 .715 .771 .756

Results represent multilevel multiple regression results estimated with Maximum Likelihood with Robust standard errors. Observations (n = 1,629) were nested in individuals (n = 543). Waves were specified on the within‐person level; girls, educational level, and ethnic background were specified on the between‐person level. All independent covariates are binary, and therefore all coefficients were standardized based on STDY‐standardization. SMU = social media use; ICC = intraclass correlation.

a

Ref. = wave 1.

b

Ref. = boys.

c

Ref. = intermediate/pre‐university.

d

Ref. = immigrant background.

e

ICC = variance between/(variance within + variance between).