Data Analysis on Safe Usage of Screen Time
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.
Motivation
There are a lot of concerns about the impact of smartphones on the mental and physical wellbeings of youth. Being the first to grow up with smartphones and social media, the young generation is bearing the unknown disadvantages while taking advantage of technological and data advancement. Personally, as a mother to two young kids, I find myself constantly trying to figure out whether to give my kids screen time, how long they should spend in front of a screen, and what kind of screen time they should have.
Data
The Youth Risk Behavior Surveillance System (YRBSS) is a nationwide survey conducted by the Centers for Disease Control and Prevention since 1991 every two years. It tracks a lot of youth risk behaviors, including computer time, among high school students. The survey reports the percentage of students who answered yes to each question among all the students who answered that question. It also gives percentages for subgroups like genders and grades.
Data analysis
In YRBSS, the description of the question about computer use is as follows: "played video or computer games or used a computer for 3 or more hours per day counting time spent on things such as Xbox, PlayStation, an iPad or other tablet, a smartphone, texting, YouTube, Instagram, Facebook, or other social media, for something that was not school work, on an average school day". This question will be referred to as "computer use". The map below shows the percentage of students who responded yes to "computer use" across the US in 2017 (the most recent year with data available).
The boxplot below shows the percentages of "computer use" among students in different grades with different genders in 2017. We can see that higher grade students spend less time on the computer while female students spend slightly more time than male students.
Grade and Gender
This question has been asked in the survey since 2009. There has been a significant increase in computer use among all grades and both genders.
Girls
It's particularly interesting that female students were using much less computer time than male students in 2009, but their computer usage rapidly increased from 2009 to 2013. At the same time, we also see an increase in unhappiness among girls.
Obesity
The obesity rate also went up at the same time.
Cyber-Bullying
While bullying among female students increased, bullying among male students dropped.
Sleep Time
While students are sleeping less in general, the trend started after 2013. So there is an offset in time between the growth in "computer use" and decrease of sleep time.
We make scatter plots of "computer use" versus these behaviors, and fit a linear correlation.
It's clear that:
- More "computer use" may make girls feel unhappier, no significant effect on boys.
- More "computer use" may increase the obesity rate.
- Increase "computer use" may be linked with increasing bullying among girls, and less bullying among boys.
- More "computer use" may not be directly linked to the decrease in sleep time.
But correlation doesn't mean causality. If we plot unhappiness versus alcohol use, there is a strong correlation as well.
Apparently, we can't conclude female students are feeling sadder because they are drinking less alcohol (can we?).
Shiny App
The shiny app I built has data from the entire survey. Users can use the interactive app to explore the data set by themselves. The source code is written in R and can be found at my GitHub page.
About Author
