Jolien Trekels

The through-line

Across everything, Trekels argues the same case: the link between screens and youth wellbeing is real, negative, and much smaller than the headlines and the politics around it suggest. The strongest available evidence, meta-analyses and large-scale studies of 100,000-plus young people such as those from Cambridge, finds a negative association between smartphone and social media use and wellbeing, but one that "verklaart ook vaak maar 1% van de welzijnsvariabelen" [3][4]. From that starting point everything else follows: if the average effect is tiny, the interesting question is not whether social media is bad but for whom, under what conditions, and through which mechanisms. Her repeated formulation is that "het one size fits all geldt niet voor sociale mediaeffecten. Dus er is geen uniform effect" [3][4].

The second half of the position is that nuance does not mean doing nothing. She is consistently sceptical of the interventions that are easiest to legislate, and consistently in favour of the ones that are harder to organise: guided use rather than withholding, collective responsibility rather than individual media literacy, and holding platforms to account rather than treating the technology as a self-contained villain [3][4]. Her own research programme runs on the same logic, moving from neural and motivational differences between adolescents [2] toward the next frontier of the same question, what competence looks like when generative AI becomes part of everyday thinking and decision-making [1].

On how big the effect actually is

The headline number is the anchor. "Er is een verband tussen smartphone gebruik en sociale media en welzijn. Dat kan gaan over depressie, angst bijvoorbeeld, maar dat verband is heel klein en het verklaart ook vaak maar 1% van de welzijnsvariabelen" [3]. She restates it without hedging in either direction: "Er is een verband, maar dat is heel klein. Het is negatief maar het is heel klein" [4]. Much of the public evidence that feels more convincing than this is, in her account, an artefact of presentation. Overlaying a trend line of teen suicide attempts on the launch date of Facebook mobile is misleading; the studies that actually link the two variables recover only a small explained percentage, and trend graphs strip out the context needed to interpret them at all [3][4].

On causality and amplification

Establishing that social media causes mental health problems is, she argues, scientifically very difficult, and the research designs available make causal claims hard to sustain [3][4]. Her working position is that social media is one factor among many and almost certainly not the primary cause, functioning more plausibly as an amplifier of vulnerabilities that already exist [3][4]. The practical corollary is blunt: "Het is niet dat als je die smartphone of die sociale media weghaalt dat dan opeens alle problemen zijn opgelost" [4].

On why some teens and not others

The organising framework here is the Swiss Cheese Model of Social Media Effects: layered buffers at the individual level (emotional maturity), the social level (close friendships and offline support) and the platform level (content and algorithms), with vulnerabilities as the holes that let negative effects through [3][4]. That is what produces positive outcomes for one adolescent and negative ones for another from the same platform. Some adolescents simply have a more socially sensitive brain and react more strongly to likes and social feedback, while strong offline friendships buffer against harm [3][4].

Her own work puts empirical weight behind exactly this interaction. Combining fMRI data, peer sociometric nominations and longitudinal survey data, she and colleagues found that neural sensitivity to low-status peers predicted greater digital status-seeking two years later, but only among adolescents with a high need for approval [2]. The finding that matters most is the negative one: neither brain responses nor social motivations alone were sufficient to predict digital status-seeking, and it was the combination of heightened neural sensitivity and stronger approval motives that did the work [2]. Social media offers countless routes to status through likes, followers and other visible metrics, and the question of why some young people chase those signals harder than others is answered at the intersection of neural and motivational difference, not by either on its own [2].

On bans and age limits

Trekels splits the ban question rather than answering it wholesale. A ban during class is scientifically defensible, because "de smartphone die aanwezig is leidt af. Daar gaat al cognitieve capaciteit gewoon in" [4]; mere presence drains attention [3]. Extending the ban to breaks goes too far. Teens connect through their phones, and marginalised or vulnerable students may lose their main source of connection and security [3][4]. Bans can also produce overcompensation after school hours driven by fear of missing out, which is itself evidence of how deeply the phone is woven into adolescent social life [3][4].

On minimum ages she is more categorical: there is no scientific basis for a fixed cutoff. "Het is niet dat je kan zeggen: oké, je bent nu 14 jaar, nu is jouw ontwikkeling compleet en kunnen we jou gerust een smartphone of sociale media geven" [3]. Emotional maturity and circumstances vary per adolescent and matter far more than calendar age; the reason a hard cutoff is attractive is that it is enforceable, which is a political virtue rather than a scientific one [3][4].

On building self-regulation through use, not abstinence

The strategy of withholding a phone until some age and then hoping the teenager arrives equipped with the right norms and self-regulation does not work [3][4]. Her line on this is the most compressed statement of her position: "Je kan alleen maar dat creëren door het hen ook te laten gebruiken" [3][4]. What the literature supports is guided use, with parental co-use, co-viewing and mediation demonstrably helping digital literacy develop [3][4].

On collective media literacy and platform accountability

Media literacy framed as something you install in individual teenagers is, in her view, the wrong unit of analysis. "Het gaat niet alleen over jongere mediawijs maken, maar het gaat ook over die platformen wel ter verantwoording houden" [4]. The full package she describes is regulation, platform accountability, parents modelling healthy digital etiquette such as no phones at the dinner table, and peer-group interventions, on the grounds that teens are highly sensitive to what their friends do [3][4].

On the platforms themselves she is careful about what can and cannot be proven. Internal documents matter: "Er zijn wel interne documenten waarin dat ook wel wordt aangehaald van: we zoeken naar manieren om gebruikers langer in de app te houden" [3]. Those documents are part of why TikTok and Snapchat settled, and features like infinite scroll are plausibly addictive [3][4]. But drawing a direct line from those design features to mental health harm remains difficult both legally and scientifically [3][4].

On science, moral panic and the media debate

She identifies a genuine split inside the field. Some scientists feed the simple story: "Het is de technologie, dat is de boeman, neem die technologie weg en we hebben die problemen niet meer. Dus die wetenschappelijke stemmen spelen daarop in" [3]. That amplification partly explains how quickly the wave of smartphone bans arrived, while other researchers insist the reality is more complex [3]. She places the current moment in a longer sequence: the moral panics around television and video games followed the same pattern, built on a hypodermic-needle assumption of uniform strong effects that science later nuanced [3]. The one genuine difference she grants is how deeply digital media is woven into everyday life [3].

On what comes next: wearables and AI

The boundary problem is about to get worse. Smart glasses, always-on wearables and always-visible AI interfaces will further blur the offline/online divide and are likely to amplify distraction and notification stress beyond what smartphones already cause, given that a nearby phone alone already consumes cognitive capacity [3][4].

Her research agenda is moving onto that terrain. The question she is now pursuing is what happens when AI becomes something people routinely turn to for thinking, navigating uncertainty and making decisions [1]. The framing deliberately goes beyond usage measurement, toward the everyday practices and competencies that develop around it: what it actually means to be AI literate, what people need to understand, question and be able to do in order to integrate AI into their lives in ways that are useful without simply deferring to it, when they reach for it, what makes them trust or challenge its responses, and how they decide when to rely on their own judgement [1].

Takeaways

• Treat the effect size as the starting point of the argument, not a footnote: the association between social media use and wellbeing is negative but "verklaart ook vaak maar 1% van de welzijnsvariabelen" [3][4].
• Distrust overlaid trend graphs, such as teen suicide attempts against social media launch dates; studies that actually link the two variables explain only a small percentage and trends lack interpretive context [3][4].
• Susceptibility is an interaction, not a main effect: neural sensitivity to low-status peers predicted digital status-seeking two years later only among adolescents with a high need for approval, and neither factor alone was sufficient [2].
• Ban phones in class, where mere presence drains cognitive capacity, but not during breaks, where marginalised students may lose their main source of connection and bans can trigger post-school overcompensation [3][4].
• Reject fixed minimum ages: emotional maturity and circumstances matter more than calendar age, and the appeal of a hard cutoff is that it is enforceable, not that it is evidence-based [3][4].
• Self-regulation is built through supervised use, with parental co-use and mediation: "Je kan alleen maar dat creëren door het hen ook te laten gebruiken" [3][4].
• Make media literacy collective, combining regulation, platform accountability, parental modelling and peer-group interventions, because "het gaat niet alleen over jongere mediawijs maken" [4].
• Internal platform documents about keeping users in the app longer, plus features like infinite scroll, strengthen the case against platforms, but the causal link to mental health harm remains hard to prove legally and scientifically [3][4].
• The next literacy question is AI: not how much people use it, but what they need to understand and question in order to use it without simply deferring to it [1].