April 18, 2017

expert reaction to study looking at toddler touch screen use and sleep

Publishing in Scientific Reports, researchers have looked at the amount of screen time use and sleeping patterns in toddlers.

Dr Amitava Banerjee, Senior Clinical Lecturer in Clinical Data Science, UCL, said:

“This is an online survey of parents regarding touchscreen device use and sleeping habits of their children. It is not clear how parents were contacted. There may well be a bias towards parents who are more worried about the effects of touchscreen use and effects of infant sleeping patterns. The survey relies on self-report of parents and gathers information at one point in time, again opening up the possibility of bias. It is not clear whether the parents or children in this study are representative of the overall population of the UK. Although the hypothesis deserves investigation, we cannot make any conclusions about the causal link between touchscreen use and sleeping patterns without further research.”

Prof. Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“I certainly wouldn’t lose any sleep over these results, if I still had young children.

“The participants in the study were hardly typical of UK families. It’s unclear whether people who choose to respond to online surveys are ever typical of the whole population, but here there’s clear evidence that they weren’t. Of the mothers involved, a huge 45% had a postgraduate qualification, and in all 86% had a university-level qualification. In the 2011 Census, only about a third of women of child-bearing age in England and Wales had a university qualification. We aren’t given much more information about the families in the survey, but if they are so untypical in this way, they could be untypical in other ways too. In particular, would parents who choose to respond to a survey about touchscreen use be typical in terms of their young children’s touchscreen use? We just can’t tell.

“It’s also important to realise that, although 715 parents participated, they didn’t all answer all the questions. Almost a quarter didn’t answer questions about their child’s sleep patterns, so the survey wasn’t as big as first appears, in terms of what’s being reported here.

“Dr Tim Smith, in his quote in the press release, makes it clear that this study does not in any way establish whether touchscreen use causes sleep problems. Taking the results at face value, babies and toddlers who spent more time using a touchscreen did tend to sleep less. But there are many possible explanations of that, besides the possibility that the touchscreen use directly affects sleep. It could even be the other way round – maybe the parents of toddlers who already sleep less are more likely to let their children use touchscreens, to keep them occupied or even with the thought that it might calm the toddlers. Or maybe there are characteristics of some families that, independently, make the toddlers use touchscreens more and also make them sleep less. Families do differ a lot, and the researchers were unable to allow for many confounding factors that might explain the differences. Apart from the age and sex of the children themselves, they took into account only the family’s TV use and the mother’s education, and mother’s education was so concentrated at higher levels in this sample that allowing for that would have little effect. So again, we just can’t tell what’s causing what here.

“The size of any effect on sleep duration is arguably not very large anyway. The study found that an additional hour of touchscreen use was linked to an average reduction of 16 minutes of sleep per day. But an hour of touchscreen use is a lot in children of this age. On average, the children in this study used a touchscreen for about 25 minutes a day. A child who used a touchscreen for this average length of time would, according to the researchers’ statistical model, sleep for about 6 minutes less on average than a child. Six minutes really isn’t much, given that on average the children in this study slept on average for twelve and a half hours a day.”

Dr Andrew Przybyslki, Experimental Psychologist and Research Fellow at the Oxford Internet Institute, University of Oxford, said:

“This a survey study of parents and their perceptions of the television and touch screen technology and sleep patterns of their young children. The work is focused on an important topic but the tentative nature of the data and the overly strong interpretations should give parents and policy makers pause. There are many reasons why the significance of the work should be interpreted with extreme caution by those who are responsible for regulating child technology use. I will detail three.

“First and foremost, the participants were told about the purpose of the study in far too specific of terms to trust the results. Parents were told the study was “to examine how use of touchscreen devices such as smart phones or tablets might influence infants’ development.” This would allow participants to guess the nature of the study and introduce a confound called a ‘demand characteristic’ (link: https://en.wikipedia.org/wiki/Demand_characteristics).  This is problematic because parents provided estimates both of their child’s screen time and their sleep behaviour. It is very likely that participants were aware, on some level, the purpose of the study, putting them into what psychologists call the ‘good participant role’ in which they able to guess and confirm the researcher’s research questions. Because the effects reported here are very small it is a leap to assume the correlations observed do not reflect this confound.

“Second, it is not clear who the sample is meant to represent or generalise to. Though 750 parents is not a small sample, participants were recruited through university participant pools and advertisements which do not reflect the general population. Indeed, parents concerned about screen time may have opted into the study at higher rates than one would expect from the general population. Though the authors, and the peer reviewers, appear to have glossed over this point, it is important to understand this is a critical shortcoming.

“Third, though modest correlations are observed it is not clear why they are related. It is possible that parent estimates of one activity (screen time) is negatively correlated with parent estimates of another activity (sleep) but it’s not clear why the two are related. The mostly likely reason is one person is providing two time estimates. The second most likely reason is that parents use touch screens to help their non-sleeping kinds wind down. The third most likely reason is that the researchers asked about a range of screen time behaviours and only included those that ‘worked’ in their presentation of the results. That understood, the paper assumes that the screen time causes the relatively lower levels of sleep.

“Fourth, there are issues in how the data are presented that hamper our understanding the real-world relations between screen time and sleep quality. Screen time measures are left out of the key correlation plot in the paper so it is not possible to know exactly how strongly sleep and screen time are related in these data or whether the relations may be nonlinear. Importantly, because of this, it is not clear that screen time displaces a key amount of sleep. Children do not need unlimited sleep and it is not clear if the effect of screen time ‘pushes’ young people over a critical threshold of some type. The argument in the paper assumes this, but critically, this evidence is not presented.

“With this understood, the work would have been more interpretable if the data were open, the hypotheses were registered in advance of data collection and the study materials were publically available. As it is, this is yet another closed-science study reporting a modest correlation between factors that confirm our fears as parents and confirm our pre-existing biases about new technologies.”

Dr Anna Joyce, Research Associate in Cognitive Developmental Psychology, Coventry University, said:

“As the first study to investigate associations between sleep and touchscreen use in infancy, this is a timely piece of research given the already controversial topic of screen use in childhood and adolescence. It shows that as early as 6 to 36 months of age, touchscreen use, which is prevalent in toddlers, is associated with shorter total sleep duration by an average of 15.6 minutes per day. This association is driven by a reduction in night-time sleep despite an increase in daytime sleep, and increased sleep latency (time to fall asleep). The increase in daytime sleep may indicate a physiological response to counteract the reduction in night-time sleep, yet children do not quite manage to catch up.

“The study benefits from a large sample and robust analytical methods with appropriate control of confounders. The use of questions from validated, existing measures is a strength; however, parent reports of children’s sleep are not always reliable, particularly for reporting sleep quality, and if the child does not share a bedroom with the parent.

“Whilst the data do not allow us to understand the mechanisms for how touchscreen use affects sleep, they do, importantly, open the way for future research to investigate these. This is an important area of research since we know that sleep is critically important for children’s healthy development; children who sleep well have better attention, memory and other cognitive abilities, and consequently do better at school relative to their peers who do not sleep well.

“It should be noted that as a correlational study, the results do not suggest that touchscreen use causes sleep loss. It could be that children who use touchscreens more frequently need less sleep, or that a third factor is responsible for the association.

“In light of these findings and what we know from previous research it may be worth parents limiting touchscreen, other media use and blue light in the hours before bedtime. Until we know more about how touchscreens affect sleep, they shouldn’t be banned completely, as there may also be cognitive benefits associated with their use, for example, as the authors showed in their previous work, for the development of fine motor skills.”

* ‘Daily touchscreen use in infants and toddlers is associated with reduced sleep and delayed sleep onset’ by Cheung et al. will be published in Scientific Reports at 14:00 UK time on Thursday 13^th April, which is also when the embargo will lift. 

Declared interests

Dr Amitava Banerjee: I have no conflicts of interest.

Prof. McConway: I have no conflicts of interest on this.

Dr Przybylski: “I have no conflicts of interest to disclose.”

Dr Joyce: “I have previously worked with and published with one of the authors of this paper (AKS), and am associated with one of the other authors (RB).”