Citation

Clinical Psychological Science 2019; 7(2): e397.

Copyright

(Copyright © 2019, Association for Psychological Science, Publisher SAGE Publishing)

DOI

10.1177/2167702618824060

PMID

unavailable

Abstract

There were errors in some of the statistics reported in this article because in some analyses, the authors inadvertently included some observations that should not have been included (i.e., data from survey forms that did not include the depressive-symptoms items). The corrections detailed below do not alter the article's results or conclusions. On page 9, in the right column, near the beginning of the first full paragraph, the correct number of degrees of freedom for the correlation between time interacting with friends and time spent on social media is 77,487. Later in that paragraph, the correct regression coefficient for in-person social interaction as a predictor of depressive symptoms is.10. Finally, at the end of that paragraph, the degrees of freedom and results for the two analyses of variance were reported incorrectly. For adolescents low in in-person social interaction, the correct degrees of freedom are 1 and 9,765, and the correct F value is 185.35; for adolescents high in in-person social interaction, the correct degrees of freedom are 1 and 11,271, and the correct F value is 2.21. The p values are unchanged. Also on page 9 in the right column, the sentence in the last paragraph that continues onto page 13 should read as follows: "The only activities that (a) predicted lower depressive symptoms and (b) decreased at d ≥.05 since 2011/2012 are in-person social interaction, print media use, and sports/exercise." On page 15, right column, Note 3, the correct number of degrees of freedom for the correlation between sports/exercise and social media use is 77,487. The correct number of degrees of freedom and correlation coefficient for the correlation between sports/exercise and in-person social interaction are 77,487 and.23, respectively. Later in that paragraph, the correct regression coefficient for in-person social interaction as a predictor of depressive symptoms is.06. Finally, the correct effect size for change in social media use between 2012 and 2015 is.20. The Supplemental Material has been revised as well All of these errors have been corrected. © The Author(s) 2019.

Language: en