The Annex: 491 families, four waves, and a link that only showed up at month 18

The Annex reads the method behind the headline. This week’s paper, in the Journal of Research on Adolescence and summarised by PsyPost, asks whether strict, punitive parenting with little warmth goes with teenagers’ beliefs that rule-breaking and aggression are acceptable, and with anxiety and depression¹.

The design. Wusirige, a PhD candidate at Syracuse University, with Bo Jian and Gabriel J. Merrin, used 491 US adolescents and one parent each from the AmeriSpeak Panel, a randomly recruited survey panel¹. Families answered four waves about six months apart between November 2021 and summer 2023; teenagers averaged about 15.3 years at the start, and 71.3% of responding parents were mothers¹. Parents rated authoritarian practices once, at wave 1, on a 0-to-4 scale; teenagers reported anxiety, depression and agreement with statements such as “Some kids deserve to be picked on” at waves 2 to 4¹. Participation fell from 413 at wave 2 to 364 at wave 4¹.

The trick. A random-intercept cross-lagged panel model separates two questions: do teenagers from stricter homes differ on average from others (between-person), and does a given teenager rise above their own usual level (within-person)¹? Headlines usually blur them.

The results. Between families, more authoritarian parenting went with modestly higher stable levels of antisocial thinking, but not with average anxiety or depression¹. Within teenagers, nothing showed up at waves 2 and 3. At wave 4, about 18 months after the parenting survey, earlier authoritarian parenting was associated with rises in anxiety and depression above each teenager’s own baseline; the depression link sat right at the threshold of significance¹. No differences between boys and girls were significant¹.

The footnotes that matter. Levels were low: the mean parenting score was 0.61 out of 4, between “never” and “rarely”¹. Parenting was measured once, so the study cannot say whether changes in parenting preceded changes in teenagers, and parent–child influence runs both ways¹. The sample was limited to cisgender, heterosexual teenagers with a biological parent responding¹. “These findings show associations, not causation,” the lead author says, and endorsing a belief is not acting on it¹.

For you. The authors’ own takeaway is modest and useful: “the absence of immediate difficulties does not necessarily mean there will be no difficulties later”¹. Explanation costs a minute more than a threat. On this evidence, the minute is not wasted, even when nothing seems wrong.