January 28, 2019

expert reaction to study on use of screen time and child development

Researchers publishing in JAMA Pediatrics argue for a directional association between screen time and child development, and recommend family media plans and management of screen time.

Dr Bob Patton, Lecturer in Clinical Psychology at the University of Surrey, said:

“This study has employed a longitudinal design to try and determine the direction of the already well established association between children’s exposure to screen based activity, and developmental outcomes, finding that increased earlier exposure was associated with lower developmental outcome scores at a later time.

“While the study does not take account of either the type/quality of screen based activity or of any non-screen activity that may also affect developmental outcomes, it does add further to the evidence base that suggests limitations on such activity could be of benefit.

“The extent to which the findings of the study may be generalisable to a broader population are less certain given to over representation of white (78%), degree educated (58%) high income households (66%) who participated in the study. However it seems likely that reducing young children’s exposure to screen based activity in favour of other more beneficial opportunities, could be advantageous.”

Prof Andrew Przybylski, Associate Professor and Director of Research at the Oxford Internet Institute, University of Oxford, said:

“In this study the researchers draw on data they collected as part of an ongoing birth cohort study based in Canada. This paper has three real strengths that should make readers take it seriously and three notable limitations which should temper how we interpret the conclusions the authors draw.

“In terms of strengths, the sample is well specified and the basic analytic approach is top notch. The researchers did their homework and properly distinguish between effects we might think are due to differences between individual children and differences in the same child over time. This provides insight into how technology impacts individual young people over time (and vis versa). This is something that most studies on screen time do not do. Second, the researchers are able to test if the correlations they observe suggest screen time variation precedes their developmental outcome or reverse. Finally, the researchers use data that policymakers might consult if they get permission (a real plus). Many studies lock up their data and this leaves the public in the dark.

“With this understood there are a few critical places where the conclusions and recommendations the authors draw are entirely premature.

“First, the fact that the correlations the researchers report are ‘statistically’ significant is meaningless and this result (statistical significance) is confused with assuming the results are practically significant. Because the sample size is large, even miniscule effects will be flagged by these models as statically significant. A true link between screen time and developmental outcomes would be a cause for worry; but increases in screen time observed here indicate between about 0.36% 0.64% of the variability in decreases in the developmental outcome. This means that upwards of 99% of the children’s developmental trajectories studied here have nothing to do with screens.

“Second, and to the authors’ credit this is noted, most of these data come from mother’s self-report of children’s screen time. We know that this a very poor way to accurately measure use and that many biases lead to both over and under reporting. It is possible, indeed likely, that some factors not measured in this study that correlate with shifts in a child’s life circumstance are driving the small correlations reported in this study. This should serve to motivate the public to demand higher (and more expensive) measures of technology use to support better basic research on this topic. 

“Finally, the conclusions drawn are overly strong for the method used. While the Ages and Stages Questionnaire (ASQ) is a fairly good questionnaire to measure a complicated thing, it is only one of many the researchers could have reported. The paper focuses on the link between screen time and this one measure. But this is only part of the picture. The study actually includes seven other measures of child developmental and includes multiple measures of physical health and sleep. There are also measures of other environmental factors such as recreation time and how people read in the house that are left out completely. Because all of these important factors are omitted we cannot put the 99% of developmental outcomes that have nothing to do with screens in context.

“To conclude, much of this study is well conducted, which is good in a research field where many studies are poorly done but there are huge limitations to be aware of in terms of the practical implications of the work. It is premature to advice that limiting screen time alone will improve developmental outcomes for children in any meaningful way.”

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

“This new research has helped a little in understanding possible associations between screen time and child development, but it’s important not to over-interpret it. I agree that the evidence from the study supports the researchers’ recommendations that health care practitioners should talk to parents about screen use, but I don’t think it gives any evidence for guidance to practitioners or parents on what might be an appropriate amount of screen use or what, exactly, amounts to ‘excessive screen use’.

“There are numerous difficulties in researching these associations. For example, how do you measure screen time, how can you take into account that different children do different things in their time using screens, and what’s the impact of changes over time in the sort of screens (TV, DVDs, tablets, computers, and so on) that children use? Very importantly though, if we want to change something about children’s development by changing the way they use screens, or indeed in any other way, we have to know what causes what. Most previous research that looked for associations between screen time and factors like children’s development or well-being has been observational and cross-sectional, that is, the researchers record data from the children and (usually) their parents, but without seeking to change anything the parents and children would be doing anyway, and they record screen time and development at the same time. Several previous studies of this kind have found that children whose screen use is high also tend to have lower scores on developmental tests (albeit, in most studies, this tendency is not particularly strong). However, a cross-sectional study can’t distinguish between, on one hand, the possibility that high screen time really does have an adverse effect on development, and on the other, the possibility that children with lower developmental scores watch screens for longer, perhaps because of their own preferences or perhaps because of the way their parents deal with developmental issues. Also, any observational study can’t rule out that the observed association isn’t causal at all, but instead that differences both in screen time and in developmental measures are separately affected by some other variables. But if the developmental issues aren’t actually caused by high screen time, then taking action to change screen time won’t change anything about the children’s development.

“The new study has helped to a certain extent in teasing out what causes what, by using data on both screen time and child development that was collected at more than one age. Such longitudinal data can look separately at associations between the amount of screen time at one age, say 24 months, and the developmental score at a later age, say 36 months, and also at associations between the developmental score at the earlier age and the screen time at the later age. This allowed the researchers to distinguish between the possibility that screen time has an association with a measurement of development at a later age, and the alternative possibility that development has an association with screen time at a later age. In very brief summary, the researchers found statistically significant evidence of the first type of association (screen time at an earlier age with development later), but not of the second type (development at an earlier age with screen time later). This is consistent with the explanation that screen time has a causal effect on development, but development does not have a causal effect on screen time, and maybe that is what’s going on.

“But the problem is that the results are consistent with other explanations too. Perhaps the association they found is still not one of cause and effect, because other variables are involved. The researchers did adjust some of their results, to allow for variables like the mother’s age and education, the child’s gender, whether the child is read to, and how much sleep the child gets. However, they were able to do that only for part of the statistical model that they used, and not directly to the parts that dealt with changes between one age and another, so some doubt about what causes what must remain. Perhaps there was still some element of developmental score affecting screen time at a later age, despite the lack of statistical significance. One of their estimates of that reverse association was actually bigger than the estimates of association in the other direction, but was not statistically significant because it could not be estimated very precisely from their data. Perhaps if they had had more data, that association would have been significant, or perhaps it would still not have been. And, as the researchers themselves are careful to point out, not all screen time is equal, and they could not take account of how children were actually using screens or look at the impact of changes in patterns of screen use over the five years of data collection.

“So this study has advanced understanding, but it certainly hasn’t established for sure that what it describes as ‘excessive screen time’ definitely harms children’s development. Their statistical analysis can’t say what might be meant by ‘excessive’ anyway. Even if there is a causal effect of screen time on development, the size of that effect doesn’t appear to be very large. Since the quantities estimated in the researchers’ statistical model relate to changes in levels of screen time and developmental scores in a rather complicated way, I can’t give a simple summary of how big the effect seems to be, but it appears to be considerably smaller than for instance, the effects of sleep time, or whether the child is read to, or family income, on development.”

Dr Max Davie, Officer for Health Promotion for the Royal College of Paediatrics and Child Health (RCPCH), said:

“This paper gives an indication of a weak association between screen use and developmental outcomes.  It is however very important to bear in mind that this paper does not demonstrate that the relationship is causal, and in fact the data shows that the association with screen time is weaker than that between developmental outcomes and good sleep, reading to the child, and maternal positivity.  

“We would, in the light of this paper, reiterate our advice that families spend time interacting as a family, that screens are not allowed to interfere with sleep, and that screen based interaction is no substitute for in person contact.”

Dr Bernadka Dubicka, Chair of the Child and Adolescent Faulty at the Royal College of Psychiatrists, said:

“This is the first study to show that increased use of screen time in very young children can be associated with slower development. It is a large, well conducted study that tracks toddlers between the ages of 2 and 5.

“These results add important weight to existing concerns that too much screen time can prevent young children from having the best start in life, by potentially reducing important opportunities for social interactions, physical activity and other experiences necessary for development.

“While this study sheds light on screen time and a child’s development, we still need more research to tell us which children are most vulnerable to the harms of screen use and the impact it may have on a child’s mental health. We also need to look at the effects of different content as there are also many positive ways of using screens.

“Parents should actively encourage their children to engage in a range of activities which promote their child’s development and give them as much face-to-face time as possible. Parents should also be aware of how much time they are spending on their screens in front of their children.”

‘Association Between Screen Time and Children’s Performance on a Developmental Screening Test’ by Sheri Madigan et al. was published in JAMA Pediatrics at 16:00 UK time on Monday 28th January.

All our previous output on this subject can be seen at this weblink: http://www.sciencemediacentre.org/tag/screen-time/

Declared interests

Dr Bob Patton: No conflicts of interest

Prof Andrew Przybylski: “I do not have any interests which might be regarded by a reasonable and objective third party as giving rise to a conflict with my role as an SMC expert in this story.”

Prof Kevin McConway: “Kevin McConway is a Trustee of the SMC.”

None others received.