Table 1 Mixed linear model results on the standnumberardized associations between screentime, number of activations, and number of notifications and hedonic/eudaimonic well-being in the << blinded for review>> study.

From: Appnome analysis reveals small or no associations between social media app-specific usage and adolescent well-being

	Positive emotion				Negative emotion				Eudaimonic activities				Eudaimonic well-being
	Coef	SE	P	VPC	Coef	SE	P	VPC	Coef	SE	P	VPC	Coef	SE	P	VPC
Screentime by App
Games	0.01	0.02	0.677	0.34	0.02	0.02	0.324	0.21	− 0.04	0.02	0.043	0.46	− 0.03	0.02	0.254	0.36
Instagram	0.00	0.02	0.980	0.33	0.02	0.03	0.549	1.00	− 0.03	0.02	0.155	0.46	− 0.02	0.02	0.423	0.36
Snapchat	0.01	0.02	0.712	0.33	0.02	0.02	0.495	0.20	− 0.02	0.02	0.474	0.45	− 0.02	0.02	0.417	0.35
TikTok	− 0.04	0.02	0.044	0.33	0.03	0.02	0.258	0.21	− 0.04	0.02	0.037	0.44	− 0.08	0.02	0.001*	0.35
WhatsApp	− 0.04	0.02	0.075	0.31	0.06	0.02	0.008*	0.19	0.00	0.02	0.882	0.45	− 0.01	0.02	0.607	0.35
YouTube	0.03	0.02	0.178	0.33	− 0.01	0.02	0.702	0.19	0.00	0.02	0.941	0.46	− 0.03	0.02	0.206	0.37
Activation by App
Games	0.00	0.03	0.907	0.17	0.00	0.03	0.903	0.99	− 0.02	0.03	0.472	0.26	− 0.04	0.03	0.108	0.21
Instagram	− 0.01	0.03	0.763	0.16	0.02	0.02	0.503	0.37	− 0.03	0.03	0.335	0.25	− 0.01	0.03	0.817	0.22
Snapchat	0.02	0.04	0.587	0.98	0.01	0.02	0.840	0.38	0.01	0.03	0.596	0.25	0.02	0.03	0.528	0.21
TikTok	− 0.03	0.03	0.238	0.17	0.01	0.02	0.712	0.39	− 0.03	0.03	0.325	0.26	− 0.02	0.03	0.429	0.21
WhatsApp	0.02	0.03	0.363	0.16	0.00	0.03	0.990	0.37	0.02	0.03	0.359	0.25	0.07	0.03	0.013	0.19
YouTube	0.03	0.03	0.211	0.17	− 0.05	0.03	0.030	0.38	0.04	0.03	0.119	0.25	0.00	0.03	0.947	0.21
Notification by App
Games	0.01	0.03	0.691	0.39	0.03	0.04	0.456	0.69	0.02	0.03	0.464	0.45	0.00	0.03	0.924	0.37
Instagram	0.04	0.02	0.053	0.37	− 0.04	0.03	0.241	1.00	0.03	0.02	0.110	0.43	0.02	0.02	0.437	0.35
Snapchat	− 0.01	0.02	0.486	0.34	0.01	0.02	0.750	0.25	− 0.03	0.02	0.144	0.42	− 0.04	0.02	0.105	0.35
TikTok	0.00	0.02	0.958	0.37	0.06	0.02	0.003*	0.26	0.00	0.02	0.827	0.43	− 0.02	0.02	0.455	0.36
WhatsApp	− 0.03	0.02	0.281	1.00	0.04	0.02	0.044	0.23	− 0.03	0.02	0.103	0.42	− 0.02	0.03	0.386	1.00
YouTube	− 0.01	0.02	0.656	0.36	0.00	0.02	0.879	0.23	0.04	0.02	0.078	0.43	0.02	0.02	0.457	0.36

Table shows the standardized relationship between well-being outcomes (Y) and the App-specific features (X) of the same day from mixed linear models, adjusting for the well-being outcome of the previous day, age, gender, and school in each model, with time (each day) as level one, and adolescents as level two. Well-being outcomes are collected from self-reported questionnaires, and App-specific features are derived using the text extraction pipeline from the cellphone screenshots of the setting pages uploaded by users, both within the same EMA study.
For example, positive emotion (y1)—screentime of Instagram (x1) is one pair of well-being outcomes – App-use feature predictors. We have 72 pairs of such outcomes (4 outcomes) and predictors (3 features per App for 6 Apps).
For each of the 72 pairs of well-being outcomes – App-use feature predictors, we fit a random intercept model and random slope model. We apply the likelihood ratio test to select the best model following the parsimonious rule. We report the estimated coefficients, standard errors, P values, and the Variance partitioning coefficients (VPC) of each model in this table.
Bonferroni correction is applied to account for multiple testing. For each well-being outcome and App-use feature, we apply an adjusted threshold for the P values. For example, regarding positive emotion with screentime, six regression models are fitted to test competing hypotheses, thus the P value threshold for statistical significance is adjusted by 0.05/6 = 0.0083. The coefficient estimates that are significant after the Bonferroni correction are marked with * and highlighted in bold.

Back to article page

Table 1 Mixed linear model results on the standnumberardized associations between screentime, number of activations, and number of notifications and hedonic/eudaimonic well-being in the << blinded for review>> study.

Search

Quick links