Lesson 10 of 10 · 20 min
Technology and the developing mind
Consider how digital technology may affect children's and adolescents' cognitive development, and draw conclusions from correlational evidence, with r, p-values, direction and third variables.
What you will be able to do
- Consider possible effects, harmful and helpful, of digital technology on children's cognitive development.
- Describe the direction and strength of a correlation from a scatterplot and r, and decide whether it is significant from p.
- Explain why a correlation does not show cause, with reverse direction and third variables.
- Draw and justify conclusions from secondary sources about technology use and an aspect of adolescent cognition.
Most Australian children now grow up with smartphones, tablets and laptops from their first years. Newspaper headlines swing between "screens are rewiring children's brains" and "worries about screens are a moral panic". As a psychology student, your job is to look past the headlines at the evidence, and to know what kind of conclusion that evidence can support.
Why psychologists ask the question. The developing brain is plastic and shaped by experience, especially in infancy and adolescence. So how children spend their time matters. Researchers have suggested several ways technology use could affect cognitive development:
- Displacement: time on screens may replace sleep, play, reading, conversation and exercise, which support development.
- Attention and executive functions: fast-paced media and constant notifications may train attention to switch rapidly, possibly at the cost of focusing on relevant information, planning and inhibiting impulses.
- Benefits: educational apps and programs, information, practice of visual and spatial skills in some games, and staying connected with friends.
Because of concerns like these, Australia's 24-hour movement guidelines recommend no sedentary screen time for children under 2, and no more than 2 hours a day of recreational screen time (not counting school work) for those aged 5 to 17.
Key idea
Most evidence here is correlational
Researchers can't randomly allocate children to years of heavy or light technology use. So most studies measure both variables (for example, daily screen time and a planning score) and look for a correlation: whether the two variables change together. A correlation can show a relationship. It cannot show that one variable causes the other.
Reading a correlation. The illustrative data below come from a made-up class investigation: 15 Year 11 students recorded their average daily recreational screen time and completed a planning task (a puzzle scored out of 20, where higher means better planning).
Illustrative data, made for teaching. Drag a point far from the line and watch r change.
| Value of r | Description |
|---|---|
| Sign of | Positive: both variables increase together. Negative: as one increases, the other decreases. |
| Size 0.75 to 1 (r from 0.75 to 1, or −0.75 to −1) | Strong |
| Size 0.5 to less than 0.75 | Moderate |
| Size 0.25 to less than 0.5 | Weak |
| Size less than 0.25 | No correlation (no linear relationship) |
For the class data, : a moderate negative correlation. Students who reported more screen time tended, on average, to score lower on the planning task, with plenty of exceptions.
Is it significant? A p-value from tells you how likely a correlation at least this strong would be in a sample this size if there were no relationship in the population. Here and . Since , the correlation is statistically significant: we reject the null hypothesis of no relationship.
Illustrative data from 8 students. The correlation is about as strong as the class data's, but with only 8 pairs it is not significant ().
A study of 100 adolescents finds , between hours of gaming and a visual attention score. Which description is best?
Why correlation doesn't show cause. Suppose the class correlation is real. There are at least three explanations:
- More screen time
- Poorer planning (difficulty organising time)
- Third variable, e.g. less sleep or family stress
- More screen time
- Poorer planning (difficulty organising time)
- Third variable, e.g. less sleep or family stress
(1) Screen time causes poorer planning. (2) The reverse: poorer planning leads to more screen time. (3) A third variable affects both. A correlational study alone cannot tell them apart.
Direction
Even if the variables are causally linked, the cause could run either way. Adolescents who find planning hard may spend more time on easy, rewarding activities like scrolling, rather than screens causing poor planning.
Third variables
A third variable could affect both. Less sleep, family stress, lower family income, or ADHD symptoms could each be linked with both more screen time and poorer planning, producing a correlation without either causing the other.
A survey finds that adolescents who use social media late at night have poorer attention scores the next day. Which is a plausible third variable?
Worked example
Interpret a correlation between screen time and planning
Illustrative data: 15 Year 11 students recorded their average daily recreational screen time and completed a planning task scored out of 20. Pearson's correlation coefficient is .
(a) Describe the relationship. (b) Use to find , and state whether the correlation is significant, given that . (c) A student concludes, "Screen time makes teenagers worse at planning." Evaluate this conclusion.
Step 1: (a) Direction and strength
is negative and its size is between 0.5 and 0.75: a moderate negative correlation. Students with more screen time tended to score lower on planning.
Step 2: (b) Calculate t
, with .
Step 3: (b) Decide on significance
, so the correlation is statistically significant: we reject the null hypothesis that there is no correlation in the population.
Step 4: (c) Name the problem
The study is correlational: neither variable was manipulated, so it can't show cause.
Step 5: (c) Give alternatives
The direction could be reversed: students who find planning hard may drift to screens. A third variable, such as less sleep, could reduce planning and go with more screen time.
Step 6: (c) Note the other limitations
Fifteen students from one class is a small, convenience sample, and screen time was self-reported, which may be inaccurate. A justified conclusion: "In this sample, more screen time was associated with lower planning scores."
Try it yourself first, then check the working one step at a time.
Answer
(a) A moderate negative correlation. (b) (); , so it is significant. (c) The conclusion claims cause from a correlation; the direction could be reversed or a third variable such as sleep could explain it, and the sample is small and self-reported. Only an association can be concluded.
Method
Describe direction and strength from .
Calculate (df = n − 2) and compare with 0.05.
Evaluate: causation, direction, third variables, sample and measurement.
Watch out for
Forgetting that a significant correlation is still only an association.
Using instead of in the formula for .
Drawing conclusions from secondary sources. A secondary source reports research that someone else did: a journal article, a review, a government report or a news story about a study. To draw a conclusion from one, ask:
- What were the variables, and how were they measured? Was screen time measured by an app, or estimated by the participant (self-report can be inaccurate)? Which aspect of cognition: planning, inhibition, flexible or abstract thinking, focusing on relevant information?
- What was the design? Correlational, or an experiment with random allocation? Measured once (a snapshot) or followed over time?
- Who took part? How many, what ages, from where? Can the result be generalised to Australian adolescents?
- How big is the relationship, and is it significant? A significant but weak correlation explains only a little of the differences between people.
- What else could explain it? Reverse direction and third variables.
- Do other studies agree? One study is never the last word.
| Study | What it did | What it found | Limits on the conclusion |
|---|---|---|---|
| Walsh et al. (2018) | Surveyed more than 4,500 US children aged 8 to 11 (the ABCD study) about screen time, sleep and exercise, and tested their cognition once. | Children who met the guideline of 2 hours or less of recreational screen time a day had better overall cognition on average. | Correlational and measured at one time; screen time was reported, not recorded; US children aged 8 to 11, not adolescents. |
| Ophir, Nass & Wagner (2009) | Compared university students who often used several media at once (heavy media multitaskers) with light multitaskers on attention tasks. | Heavy multitaskers were more distracted by irrelevant information and slower to switch between tasks. | Groups were not randomly allocated; young adults, not adolescents; later replications have been mixed. |
Worked example
Draw a conclusion from a secondary source
A news website reports: "Phones are wrecking teens' focus. A new survey of 1,200 Australian 14- to 16-year-olds found that those who check their phones most often during homework scored lowest on a test of ignoring distractions (, )."
Using this secondary source, draw a conclusion about the relationship between phone use and adolescents' ability to focus on relevant information. Justify your conclusion and suggest how the evidence could be strengthened.
Step 1: Identify the variables and design
Co-variables: how often students check their phones during homework, and a score on a test of ignoring distractions (focusing on relevant information). Both were measured, not manipulated: a correlational survey.
Step 2: Describe the relationship
: a weak negative correlation (its size is between 0.25 and 0.5). With 1,200 participants, , so it is significant: unlikely to be chance, but small. Phone checking accounts for only a small part of the differences in focus between students.
Step 3: Consider other explanations
Reverse direction: students who already find it hard to ignore distractions may check their phones more. Third variables: tiredness, stress or ADHD symptoms could affect both. Phone checking was probably self-reported.
Step 4: State the conclusion
"Among these Australian 14- to 16-year-olds, checking a phone more often during homework was weakly associated with a poorer ability to ignore distractions." The headline's "wrecking" claims a cause and a large effect that the evidence doesn't support.
Step 5: Suggest how to strengthen it
Record phone use objectively with an app; measure the same students over time (does heavy checking come before a decline?); control for sleep and stress; or run an experiment in which students are randomly allocated to do homework with their phone in another room or on the desk.
Try it yourself first, then check the working one step at a time.
Answer
There is a weak but significant negative association between phone checking during homework and the ability to ignore distractions in this large sample. Because the study is correlational, the direction could be reversed or a third variable could explain it, so the evidence does not show that phones cause poorer focus. Objective measures, following students over time, controlling third variables or an experiment with random allocation would strengthen it.
Method
Identify the variables, how they were measured, and the design.
Interpret (direction and strength) and together with .
Consider reverse direction and third variables.
Write the conclusion as an association; suggest better evidence.
Watch out for
Treating "significant" as "large". With 1,200 participants a weak correlation can be significant.
Repeating the headline's causal language.
Classify the variables in a correlational study of screen time and planning, and see why it has co-variables rather than an IV and a DV.
Exam tip
Writing a conclusion from correlational evidence
A strong answer: (1) states the direction and strength of the relationship with the value of ; (2) says whether it is significant, with compared to 0.05; (3) uses "is associated with", never "causes"; (4) names a likely third variable or the reverse direction; (5) notes a limitation of the evidence (sample, self-report, one-off measurement) and what would strengthen it.
Common mistake
"Linked to" in a headline is not "causes"
News stories often turn "screen time is associated with lower scores" into "screens damage children's brains". When you use a secondary source, report what the study actually measured and found, and keep the language of correlation.
Summary
Technology may affect cognitive development by displacing sleep, play and reading, and by shaping attention and executive functions; it may also bring benefits.
Most evidence is correlational: technology use and cognition are measured, not manipulated.
Describe a correlation by its direction and strength (); it is significant when . Significance depends on sample size as well as .
A correlation cannot show cause: the direction could be reversed, or a third variable could affect both.
To draw conclusions from secondary sources, check the variables and measures, design, sample, size of the effect, alternative explanations and agreement with other studies.
Finished the lesson?
Try the 2 quick checks above and the lesson completes itself, or mark it complete now.