We’re in an era when countries like Australia and the U.K. are banning social media for kids under 16, most U.S. states are keeping phones out of schools “from bell to bell,” and parents are figuring out just how much time online is good for kids. Here’s a new study on how social media use predicts worse grades in school.
The research was summarized in the UK newspaper The Guardian, and the original work appeared in Nature Human Behaviour. This is an opportunity to apply what you’ve learned in Chapter 9.
Here are some excerpts from The Guardian coverage:
Children who open social media accounts at the start of secondary school appear to perform worse in tests than those who delay, with researchers suspecting that the constant checking for updates is a major distraction for many students.
Pupils who opened their first social media account aged 11 to 12 scored lower in mathematics and reading tests than those who waited a few years, with scores at age 16 differing between the groups by the equivalent of about six months of learning, the study found.
The work does not prove social media is to blame, but the researchers took account of many other factors that could skew the results, such as the students’ academic performance before they had social media, their family structure and parents’ education.[…]
Gui and his colleagues analysed surveys on the social media habits of more than 5,000 Italian schoolchildren and their results in standardised maths, Italian and English exams at age 16. When tested [later on], those who had opened their first social media account at 11 or 12 scored worse than those who waited until at least 14 years old. [The pattern held for Italian exams and math exams, but not English.]
a) The language here suggests that these authors used multiple regression to study their question. In one of the analyses, the dependent (criterion) variable was Italian exam score at age 16. What were the independent/predictor variables? (Hint: There are four–the main predictor of interest and three others).
b) See if you can sketch a little regression table. Put the DV (Italian scores at age 16) at the top, and list the four predictor variables underneath. Think about each beta–should it be positive or negative?
c) Do the same for the other DV, Math scores. Put the DV (math scores at age 16) at the top, and list the four predictor variables underneath. Think about each beta–should it be positive or negative?
d) The authors find that a later age starting social media is associated with a higher math score at age 16. Imagine a critic brings up an internal validity problem — specifically, maybe kids who did better in school at age 11 or 12 were less likely to use social media when young and second those better students were also more likely to do well in school at age 16. In the present study, such a critique is not valid. Why?
e) Multiple regression can control for some internal validity issues statistically. In this case, the authors controlled for school achievement at age 11 or 12, family structure, and parental education. Think of another potential internal validity threat that they could have controlled for, but did not.
f) This was a correlational study because all the variables were measured. Does it have temporal precedence? Why or why not?
g) The authors of the study proposed a mediator for the relationship between early social media use and later school performance. As you read the quote below, sketch a mediator diagram of what it is describing (similar to Figure 9.12).
“We speculate that one of the mechanisms is the continuous state of alert that social media used on a smartphone exerts on minors’ lives. The continuous need to check messages, to see if something is happening in the online world, why you are engaged in something in the physical world you have around you,” Gui said. “This means being less present in the physical space of the family or the house. And from a cognitive point of view, it could mean being less engaged in something that takes time, as homework can for a student of that age.”