Reply to: Underestimating digital media harm
[METADATA ONLY; THE REPLY IS PAYWALLED AND WAS NOT READ. Matters Arising, so no abstract exists by article type. Author order confirmed at api.crossref.org; two authors, Orben first. Code and data deposited at OSF, 10.17605/OSF.IO/BYQM5 and OSF.IO/7XHA2. Everything below about substance is inferred from the verified reference list and is flagged as inference rather than argument text] Orben, A., & Przybylski, A. K. (2020). 'Reply to: Underestimating digital media harm.' Nature Human Behaviour, 4, 349-351. doi:10.1038/s41562-020-0840-y. NOT A STUDY: a published reply to the 2020 comment, in the same journal. THE REPLY'S BODY WAS NOT READ AND THIS BATCH HOLDS NO SENTENCE OF ITS ARGUMENT. What the verified reference list shows is that it answers on methods ground, citing Simonsohn on specification-curve analysis, Silberzahn and colleagues 'Many analysts, one data set', and Altman and Royston 'The cost of dichotomising continuous variables'. The last is the sharpest available clue to its line, that splitting a continuous exposure into categories, which is what a heavy-users-versus-non-users comparison does, inflates apparent differences. That clue is INFERENCE from a reference list and must be labelled as such wherever it is used. DO NOT PRINT the widely repeated claim that the reply re-ran the specification curve and still found roughly a fraction of one per cent of variance: it could NOT be verified at source. THE FACT WORTH CARRYING INSTEAD, which is verified: the exchange happened inside the journal, under both names, with code and data deposited publicly. This is a documented methodological dispute with a published reply and open code, not two camps shouting past each other, and it must not be written as a feud.
[METADATA ONLY; THE REPLY IS PAYWALLED AND WAS NOT READ. Matters Arising, so no abstract exists by article type. Author order confirmed at api.crossref.org; two authors, Orben first. Code and data deposited at OSF, 10.17605/OSF.IO/BYQM5 and OSF.IO/7XHA2. Everything below about substance is inferred from the verified reference list and is flagged as inference rather than argument text] Orben, A., & Przybylski, A. K. (2020). 'Reply to: Underestimating digital media harm.' Nature Human Behaviour, 4, 349-351. doi:10.1038/s41562-020-0840-y. NOT A STUDY: a published reply to the 2020 comment, in the same journal. THE REPLY'S BODY WAS NOT READ AND THIS BATCH HOLDS NO SENTENCE OF ITS ARGUMENT. What the verified reference list shows is that it answers on methods ground, citing Simonsohn on specification-curve analysis, Silberzahn and colleagues 'Many analysts, one data set', and Altman and Royston 'The cost of dichotomising continuous variables'. The last is the sharpest available clue to its line, that splitting a continuous exposure into categories, which is what a heavy-users-versus-non-users comparison does, inflates apparent differences. That clue is INFERENCE from a reference list and must be labelled as such wherever it is used. DO NOT PRINT the widely repeated claim that the reply re-ran the specification curve and still found roughly a fraction of one per cent of variance: it could NOT be verified at source. THE FACT WORTH CARRYING INSTEAD, which is verified: the exchange happened inside the journal, under both names, with code and data deposited publicly. This is a documented methodological dispute with a published reply and open code, not two camps shouting past each other, and it must not be written as a feud.
Written from the mETADATA ONLY; THE REPLY IS PAYWALLED AND WAS NOT READ. Matters Arising, so no abstract exists by article type. Author order confirmed at api.crossref.org; two authors, Orben first. Code and data deposited at OSF, 10.17605/OSF.IO/BYQM5 and OSF.IO/7XHA2. Everything below about substance is inferred from the verified reference list and is flagged as inference rather than argument text.
What our sources record
Recorded when this source was checked, figures included, with page references and quotation marks removed.
Orben, A., & Przybylski, A. K. (2020). Reply to: Underestimating digital media harm. Nature Human Behaviour, 4, 349-351. doi:10.1038/s41562-020-0840-y