Most young Australians intended to continue using social media, even if the ban came into force

This study is interesting in that it seeks to present an analysis from a longitudinal perspective and to examine, empirically, the fundamental assumptions underpinning Australia’s Social Media Minimum Age Restriction (SMMAR). However, it is important to note that its methodological limitations (particularly the relatively small sample size, the homogeneous socio-demographic composition limited to urban and affluent participants, and the self-reported nature of the data) restrict its generalisability. Self-reporting is valuable as it provides insight into adolescents’ intentions, but it inevitably introduces inherent biases stemming from social desirability that may not accurately reflect actual behaviour. These limitations do not invalidate the study, but they do suggest that its findings should be interpreted as preliminary indicators rather than conclusive evidence. The study uses the generic term ‘social media’, without distinguishing between TikTok, Instagram, Snapchat, YouTube, Discord, etc. Each platform has very different dynamics, risks and social functions, so a more detailed analysis would be of interest.

What is interesting about this study is the analysis of the effectiveness of age-restriction policies and the stated intentions of adolescents, as it helps us understand how they experience these restriction policies. It appears that 75 per cent of teenagers planned to continue using social media by finding ways to circumvent the restrictions. This study joins others that are beginning to be published which suggest something interesting: that simply raising the minimum age for accessing social media is not enough to resolve teenagers’ wellbeing issues. In other words, it seems that legal restrictions alone are not sufficient and that additional strategies are needed for them to be effective. It is also interesting that the motivation to continue using social media is not primarily driven by mental health issues, but by other reasons (social connection, identity, entertainment).

Another important point acknowledged in the study itself is that it only included urban and affluent participants, thus excluding the perspectives of adolescents in rural or isolated areas. In these areas, social media could be important tools for connecting with others, accessing information or various educational opportunities. For vulnerable and geographically dispersed populations, restricting access to social media may have different social consequences. Therefore, whilst this study is interesting and contributes to the debate on digital policies, it cannot be applied across the board.

Finally, it is worth noting that longitudinal studies bring us closer to understanding causal relationships than simple cross-sectional studies, so it will be interesting to follow the results of the authors’ research closely.

Is the study of good quality? Are the conclusions supported by sound data?

“The study is methodologically very simple. In the first part, the study examines Australian adolescents shortly before the ban on social media use by under-16s came into force:

• a) A very weak or negligible association between having a social media account and a higher score on a standardised anxiety scale, and between having a social media account or having ever used social media and the score on a depression scale.
• b) A weak relationship between having ever used social media and anxiety scores, and between an earlier age of onset and higher scores on both depression and anxiety scales.

The assessment of these effects as weak or very weak is based on standard criteria for the social sciences, although these remain arbitrary thresholds. More important is the assessment of these effects in terms of clinical relevance. None of the regression coefficients is of sufficient magnitude to be considered a clinically significant effect.

Furthermore, the link describing further methodological details is not yet operational, so I have been unable to carry out a careful assessment of all of them. Based on the article’s description, these analyses share the same interpretative limitations as similar studies. Firstly, they are based exclusively on self-report measures and, secondly, the age of onset of use is assessed retrospectively, which implies possible recall bias or reliability issues.

No information is provided regarding the possible control for confounding variables, although such information may be available in the supplementary material, which is not accessible.

The second part assesses the intention to continue using social media following the ban, and it is noted that the majority of adolescents surveyed intended to continue using them. Once again, this appears to be a rather simple analysis, with two significant limitations. Firstly, intention is not a direct measure of actual subsequent behaviour; and secondly, based on the available information, the sample is not representative, meaning that the observed percentages of intention cannot be extrapolated to the entire adolescent population.”

How does this study fit in with the existing evidence?

“The first part is perfectly consistent with the finding that social media usage variables correlate very weakly with indicators of mental health. Previous meta-analyses and systematic reviews have observed, beyond the data from this specific study, that these correlations:

• a) are all the weaker the greater the methodological and statistical rigour of the study,
• b) are partially explained by third variables.”

What are the implications for the real world?

“The implications are very limited. The discussion is particularly weak in this regard, as it completely ignores the inherent limitations of this type of study and overlooks the necessary discussion regarding the size and relevance of the reported effects. The assertion that ‘Overall, these findings agree that early adolescence is a period of vulnerability to the adverse effects of social media on mental health, and support the argument in favour of delaying adolescents’ access to these platforms’ is not at all supported by the results.

The finding that a significant proportion of adolescents have no intention of complying with the restrictions on social media use laid down by law is interesting, and some analyses had already predicted this as one of the major weaknesses of this type of blanket ban. The provisional data available now that the law is in force point in the same direction, namely that the ban has not been successful—or at least not to the extent expected by its proponents—in its aim of substantially reducing social media use by adolescents.”

Is the study of good quality?

“We can say that this is a well-designed study, but one that measures very little. The study design, research question, sample, etc., are all clearly defined. However, we can raise several objections to the results: firstly, the analysis presented is cross-sectional, so we cannot establish a cause-and-effect relationship, only associations (do social media cause anxiety, or do teenagers turn to social media because of their anxiety?). Secondly, and more importantly, the effect size is very small, as social media accounts for only between 1 per cent and 5 per cent of mental health in this age group. Thirdly, and related to the above, it is precisely what the law prohibits that appears in the study as being least associated with distress.”

How does this study fit in with the existing evidence?

“It fits perfectly with the existing evidence, but this is not a positive thing. Indeed, one of the points most frequently raised by researchers who are critical of the moral panic surrounding this issue is that there are a large number of studies attempting to demonstrate the harmful effects of social media on children’s mental health using cross-sectional analyses (which do not prove causality) and with very small effect sizes, as is the case here. Generally speaking, all these studies conclude that an association between social media and mental health exists, but it is very small.

However, the analysis of users’ intentions following a ban is interesting. But this also confirms what we already knew: despite bans, teenagers will continue to use social media.”

Have the authors taken confounding factors into account? Are there any significant limitations that need to be considered?

“In this respect, the study is also consistent with a vast number of previous works on the subject: it does not control for key socio-demographic variables which disciplines such as sociology have already shown to have a much greater impact on young people’s mental health. Socio-economic status, family structure and dynamics, support networks, etc., are factors that have not been taken into account in the analysis and which account for a much greater variation in mental health than social media, the effect of which, as we have said, is very minor. What’s more, the sample itself focuses on urban boys and girls from affluent families, who are not exactly the group most at risk.”

What are the implications for the real world?

“Undoubtedly, this is something we have been discussing ever since the measure was introduced in Australia, as we outlined in this article: that measures banning social media are ineffective, as young people will continue to use them, but now with less oversight, given that they can circumvent the restrictions. It is essential to focus on the platforms’ actual obligations regarding content moderation, transparency of their algorithms, and preventing harmful passive consumption such as endless scrolling; and, ultimately, their responsibility for protecting children. Furthermore, supporting families and promoting critical digital literacy from an early age must be the primary focus of these measures. A ban, as well as being ineffective, may have the opposite effect: creating less safe spaces for children and failing to protect the vulnerable groups who use these platforms as a support network.”