The Wonderful World of Stats

Beware of Television

It not only destroys the mind, it can kill a child.



Now we are talking about 180 deaths per year, each one a tragedy.

A similar report from the U.S. Consumer Product Safety Commission last year estimated 42,700 injuries and 180 deaths associated with appliance, furniture and television instability and tip-overs from 2000 to 2006; 87 of the deaths involved televisions. The number rose from seven in 2000 to 23 in 2006.
abcnews.go.com/Technology/wirestory

It was unclear from the article whether the "100 emergency rooms" were nation-wide or State of Ohio only.

It is unclear why 1990 was chosen as the initial year. It is also unclear why they think there is a trend. The eyeball thinks it looks like a step shift in 2000, no doubt due to W. being elected president that year. That is, it was relatively flat, a stationary series, prior to 2000 and likewise flat from 2000 on, but with different medians.

How can you tell a shift from a trend? Find the median of all the data shown. It's hard to read off the chart, but it seems as if all eight points from 2000 on are above the median, while eight of the nine points before 2000 are below the median. (The ninth point in '96 seems to be the median.) If it were a trend, there would be a steady increase in the number of points above the median if we broke the chart into quarters.

A trend means that a cause was operating continuously throughout a period of time. A shift means that a cause operated at a particular point in time and has continued since. This suggests that something happened in 2000 (or in 1999) to increase the number of reported falling furniture injuries.

Perhaps a difference in the reporting methods or a greater awareness of focus on the problem? Always check the measurement system first. I don't know the answer. I do know the article is overwrought. In 2006 the Statistical Abstracts of the US, 2009, Table #117 (a pdf file) counts 3868 deaths "by accident" for children 4 and under. It is not clear that the 23 deaths ascribed to TV in that year makes it the number one priority in reducing child deaths-by-accident, not to mention all child deaths.

The media gets very excited over small numbers of exotic deaths. Remember the panic a number of years back about Shark Attacks? The Year of the Shark sorta disappeared after 9/11, but each month there were more and more news stories about sharks. This, for a total of about four actual attacks. This article suggests your child would be in more danger watching Jaws on TV than swimming in the ocean.