The Impact of Television: A Natural Experiment in Three Communities

2 posts

Lone Wolf
The Impact of Television: A Natural Experiment in Three Communities

By Tannis MacBeth Williams

Television is so awful, and kids watch so much of it, that it's not surprising that there's been much sociological research on its effects. Most of the studies, though, suffer from the third-factor effect - yes, kids who watch a lot of TV tend to be more aggressive, but could a third factor explain both behaviors? Kids with lower IQ scores also tend to watch more TV and be more violent, as do kids from poorer families. The correlations between heavy TV watching and other behaviors could be merely effects from common causes.

So how can the effects of television be disentangled from all others? This book, "The Impact of Television" exploited a unique situation to do so. There was a town in central British Columbia that could not get TV because it was situated in a remote valley. BC is quite mountainous and TV signals don't carry far. In 1973 the town elders convinced the Canadian Broadcasting Corporation (the CBC, the official channel) to install a transmitter just for them. The town would get hit with television not in its early formative stage, but in its mature and virulent form. Tannis MacBeth Williams, a professor of sociology at the University of British Columbia in Vancouver, heard about the installation, and decided to test the hypotheses about the effects of television by looking at the town before and after TV arrived.

Now, hard science types routinely denigrate sociology for its lack of controlled experiments, ignoring the fact that all field sciences have the same problem. Astronomers don't get to try out different galactic structures in the lab, nor do geologists get to bang continents together to see what happens. Like sociologists, they have to find the reasons for things from studying them as they are. In this case, however, there was a beautiful example of a community where the effect of a change in one variable could be observed.

Williams and twelve other faculty and students from UBC did extensive surveys around the town, which she code-named Notel. They also did control studies at a town with a single TV channel (Unitel), and one with four broadcast channels and cable (Multitel). They studied the towns twice: once before TV came to Notel (Phase 1), and again after it had had TV for two years (Phase 2). The towns were all similar in size (around 700 within town limits, and several thousand in the surrounding district), economic base (logging, mining, and farming), income (around C$7000 per family), and class structure (~10% professional/business, ~60% skilled labor/farmers, ~20-30% unskilled labor). All had small libraries, weekly newspapers, telephones, and regular road and rail service.

Using towns with and without TV allowed Williams to filter out changes over the two years that might be common to all rural towns in BC. Using towns with only one channel versus many decreased the effect that the programming itself had. Notel and Unitel got only the CBC, whereas Multitel got the CBC and the US networks. The CBC does run commercials, but it had a 60% Canadian content requirement and did not run ads on shows for kids under 12. (Interestingly enough, many Canadian cities get the US public network, PBS, and in fact Canadians contribute a good part of its funding.)

The studies were done on various grades in the public schools, and on the towns as a whole. They were both longitudinal and cross-sectional, that is, if students in grade 2 were studied in Phase 1, the same students (if they could be found) were studied in grade 4 two years later, and new students in grade 2 were also studied. Background data on the students was also gathered, such as their IQ scores (which, incidentally, are illegal to give out in California) and the socio-economic status of their families.

Reading the book, I was impressed by the study's detail and rigor. It looks like model work, and the methodology alone has interesting lessons. Thomas Gilovich, a Cornell professor of sociology, noted in his book "How We Know What Isn't So" that social scientists are far more aware of and able to deal with the inherent randomness of phenomena than are scientists in deterministic disciplines like, say, chemistry. Sociologists know about the human tendency to find order in randomness, to extrapolate from limited data, and to create cause and effect from merely related events. In fact, the skeptical attitude induced by the probabilistic sciences is much more useful to the problems of everyday life and of citizenship than the dogmatism of the deterministic sciences. He thought, then, that sociology is a more useful (and probably more interesting!) science to teach than chemistry.

So what did Williams discover? She set up a number of specific hypotheses and devised experiments for each. I'll go through them in order of least effect seen to most seen in an unsubtle attempt to keep you reading:

Sex Roles

Television is a strongly sex-typed medium. Most of the characters in programs are male (75% by some studies), and few women are shown as employed (or at least were then). This study measured how different children thought males and females were. Each child was given a list of characteristics (e.g. hard-working, honest) and asked whether they were more typical of boys or girls. The questionaire also asked if they were more typical of fathers or mothers. A lot of differences would indicated strong sex-typing.

Television seemed to increase the differences that kids saw between the sexes, but mainly between boys and girls. It seemed to have no effect on how children perceived adult roles.

Adult Creativity

This was measured by giving a set of adults two problems: the Duncker Candle Problem and the Nine Dot Problem. For the first, the testee is given a vertical piece of cardboard, a candle, a box of tacks and a book of matches, and asked to attach the candle to the cardboard so that it can burn freely. The expected answer is to empty the box, tack it to the cardboard, and set the candle in it. A higher proportion of Notel adults got this than Unitel or Multitel, and they got it faster. In the Nine Dot problem, people are asked to connect the nine dots in a 3 x 3 array with only four lines and without lifting the pencil. Too few people actually got this one to make any estimates, unfortunately.

Leisure Activities

It's hard to find out what people actually do with their time. The amount of time people actually spend watching TV as measured by, say, the Nielsen diaries, is notoriously inaccurate, as Nielsen now admits. Even so, Williams wanted to learn if TV affected all the other things people did. She sent out questionaires all over the town that listed all the public places (e.g. parks, libraries) and public events (e.g. sports, dances, churches) and asked people if they had been there in the preceding year. As a check, she also included activities that did not exist, but found that people were honest.

The biggest effect was on sports - the number of sports events dropped in half in Notel. There was little effect on religious, civic, or business activities. The group most affected was those over 55; their participation in public events dropped dramatically, probably for the obvious reasons. The authors theorized that this could cause a widening gap between young and old, simply because the old were rarely seen any more. Overall, the number of activities did not change much, but the amount of participation did.

Child Cognition

The main test here is called Alternate Uses, where the kids are given five objects (a magazine, a knife, a shoe, a button, and a key) and asked to write down all the different uses they could think of for them. There's no time limit. They tested grades 4 and 7 both times, and 6 and 9 in Phase 2 (as part of the longitudinal study). Interestingly enough, there was little correlation between IQ and creativity. They also attempted to look for originality (uses that no one else thought of) as opposed to sheer number, but originality and number were so correlated that they weren't separable.

There was a noticeable drop in this after TV came. The raw scores dropped from 32 in Notel Phase 1 to the low 20s with TV, a drop of about 40%. However, there was little effect on vocabulary, visio-spatial skills, and only a small drop in IQ scores.


Reading fluency was measured by a tachistoscope, which flashes a word on a screen for a few milliseconds. The quicker one can recognize it, the more skilled a reader you are. Fluency was higher in Notel in grades 2 and 3 in Phase 1, but dropped back to similar levels with Unitel and Multitel in Phase 2. However, there seemed to be little effect on fluency among 8th graders, possibly because their reading skills had been established by that age. The researchers thought that TV hindered the acquisition of reading skills simply because it took up so much time. Reading is a difficult skill, after all, and doesn't become pleasurable or useful until you're good at it. If TV distracts you from practising it in that crucial stage, you may never get good reading habits. Once you have it, though, TV seems less harmful.

An interesting effect was hypothesized by the researchers. TV was originally thought to be a great leveller, since both rich and poor would be watching the same programs and would be provided with the same news and information. Instead, the reverse may be happening. Increased TV viewing in poorer households harms those kids' reading abilities and habits. They thus do less reading throughout school and so have less access to the primary (although far from the only) source of knowledge. The gap in education is thus widened, not narrowed.


This was the largest effect. Second graders were twice as aggressive towards each other (as measured by acts of pushing or taunting) after TV as before. The method here was to introduce two observers to the classroom and playground. After a week, when the children were used to their presence, they would record the aggressive acts that they saw. Each would pick a child at random and observe him/her for a minute. The observers would occasionally focus on the same child in order to check each other. They watched second graders because older kids tend to just hang out in groups rather than actively playing. Boys were more physically aggressive than girls (surprise), but they were equally verbally aggressive.

Although most aggressive acts were committed by a very few kids, the overall level rose for all children, and for both boys and girls. The levels also rose in Unitel and Multitel, but by nowhere near as much. Kids who were not aggressive before TV increased as much as kids who were. The researchers wondered if there had been some change in playground policy in the two years, but the schools all had the same principals, and they all had the same basic policy: only intervene if it looks like someone will get hurt. Such a policy wouldn't affect verbal aggression in any case.


To sum up: the introduction of television made kids more aggressive, harmed the acquistion of reading skills, decreased creativity scores, and cut participation in non-TV leisure activities. The important question then becomes: are these effects general or are they restricted to small-town British Columbia?

The book ends with a discussion of this. Williams believes the results are generalizable for the following reasons:

① The general test scores of children here match those in the rest of Canada, and overall demographics are similar.

② The effects were much more pronounced between Notel and Unitel/Multitel than between Unitel and Multitel. That is, the presence of TV itself was the important factor, and not its content.

③ The side conclusions of the studies, such as the lack of correlation between IQ and creativity, match those of other studies. That is, the administration of the tests is not unique to this study.

④ The conclusions with respect to television match those of other studies.

These conclusions are not novel and are not peculiar. The effects of TV are obviously great and are obviously far-reaching. What all the effort, all the rigor, all the detailed analysis of this study shows, is that the effects are negative.
Oh noes