This summary is intended as a supplement for the Comm 80 course. It provides a general overview of the course concepts but omits crucial details discussed in class. You should always attend class, take good notes, and study a variety of sources for maximum learning.

Academic Achievement

Past research on the effect of media use on academic achievement shows a mixed record. Outcomes typically fall into one of three categories: no effect, a negative effect (more media, worse achievement), or positive effect (more media, better achievement). Clearly the relationship is not a simple one. Two variables appear to make things more complicated. First, it depends upon how you measure media use (total TV per week, type of programs watched). Second, it matters how you measure achievement (GPA, IQ, standardized test).

Potter's Research

Consider Potter's study where these two measurement issues were controlled. Potter conducted a survey of over 500 8th and 12th graders. The students completed a self report form which measured their total time watching TV and also measures of specific programing (cartoons, MTV, sports, etc.) Potter then got the students' scores on standardized achievement tests. (You remember taking those: all day long, finger cramps, time pressure.) He then linked these two sets of scores together as "predictors" (media use) with "outcomes" (achievement on reading, math, science, and social science). Very straightforward design here: How is media use related to several global measures of academic achievement?

The results were interesting and demonstrate why you need to look at this relationship in a fine-grained way. Consider these four main findings. First, Potter found that type of programming made a difference. With "bad" programs (cartoons, MTV, sports, soap operas, late night TV) there was a small negative effect on achievement as measured by the standardized tests. By contrast with "good" programs (PBS, documentaries, news) there was a small positive effect.

Second, Potter found time displacement effects. He divided the students into three samples depending upon the amount of total time per week watching TV: Low (0-10hrs), Moderate (10-30hrs), and High (30+hrs). Within the Low and Moderate groups, Potter found a small positive effect for TV. However, within the High exposure group, there was a negative effect for TV. The following table illustrates this finding.

Low Exp (0-10 hr/wk)Mod Exp (10-30hrs/wk)High Exp (30 plus hrs/wk)
more TV better Ach
more TV better Ach
more TV worse Ach

Third, Potter discovered the "Late Night" effect. Not too surprisingly, students who reported watching the most late night TV (regardless of anything else they watched) also had worse achievement and watched more total TV.

Fourth, Potter reported the average effect in an Effect Size Window (check you workbook here). He found a small effect (45/55, right?) such that more total time with TV was associated with worse achievement. Now, most of what Potter found tends to demonstrate a negative effect, but there's a logical problem here. Because of the survey methodology, we do not have "causal" data. The data here are correlational and support both of the following statements. One, students who watch more TV (and more "bad" TV) do worse in school. Two, students who do poorly in school watch more TV. Do you see the problem? We're stuck with the dilemma of which came first, the chicken or the egg? To solve this problem and determine whether watching more TV can cause worse achievement, we need an experiment.

The "Natural" Experiment with Tannis MacBeth Williams

In the early 1970s Williams read about a small Canadian community that had never had TV (geography blocked reception), but that was soon going to get cable technology. She had a flash of insight. This small town was a "natural" experiment. Just by dumb luck, a community that had never had TV was going to be "assigned" to receive it. Now while this is not a true experiment (where the researcher does the assignment), it still functions like an experiment.

Here's what Williams did. She found two other nearby communities that were very similar to the TV-less town and decided to use them as control. She called the first community, Notel, and the two other control communities, Unitel and Multitel. (Now why did she use those names?) The communities had small populations of roughly 600-700 residents that were just typical small towns with schools, churches, and Little League. Just that one crucial difference concerning TV.

Williams then created a nice design. She decided to track these three towns over time. She and her research team visited each town three times --- in 1973, 1975, and 1977. Now, in 1973, Notel still did not have TV, but later that year cable arrived and so did television. So, Notel starts without TV, then for 1975 and 1977, television is part of the community. And, of course, TV was always available in Unitel and Multitel. Here's the design:

Time Period

During each measurement period, Williams and her team observed a lot of different variables in the communities. Right now, we only concerned with their research on academic achievement. In all three towns and during all three measurement periods, Williams measured three different indicators of achievement: Reading skill, creativity, and problem-solving/persistance.

Okay, get the big picture. Three towns, two always have TV, while one starts without TV, but then gets it. Three different indicators of learning are measured during three different measurement periods. We can make two interesting comparisons here. First, compare Notel to the other communities at any one time. (Is Notel different from the others at baseline in 1973 before they get TV?) Second, compare Notel to itself over time. (Is Notel different in 1973 before TV compared to Notel in 1975 after TV?)

What happened? Two main points. First, Unitel and Multitel, our two control towns, showed identical scores during all three measurement periods, so to simplify reporting we'll collapse those two groups into one and call the new group, Alltel. Second, the general pattern of results for all three achievement variables (reading, creativity, and problem-solving) were very similar: Notel before TV (1973) started with better scores compared to Alltell, but by the end of the experiment, Notel got worse until its performance was identical to Alltel. (See the following Figure.)

The effect sizes varied with the three different measures, ranging from large effects for reading to small effects for problem solving. But, again, the pattern was always the same: Notel started off better, but gave away that superior performance after TV came to town.

Thinking and Watching TV with Blake Armstrong

So far it is clear that TV use is not helping academic achievement, but why? What is going on in our use of TV that produces this negative effect? Blake Armstrong posed that question and then designed a cool experiment to test the answer. Here's what happened.

Blake conceptualized the problem this way. When you try to study with TV you are doing two things simultaneously. This is called "dual task processing" for obvious reasons. When you engage in dual task processing the two tasks could help each other (facilitation), could disrupt each other (interference), or be independent of each other (no effect). Okay, when you study with TV what happens? Blake solicited college adults at Michigan State to participate in an experiment. (This was his dissertation research at MSU.) If you had been a volunteer, here's what would have happened. You would have shown up at the Comm Studies Lab and been taken to a room that looked a lot like an apartment. It had a sofa, chairs, a table, and, over there, a TV. You'd sit down and Blake would tell you, "Here's the deal, I'm gonna you some tests and I want you to do the best you can on them." The tests measured these tasks: short term memory, problem solving, creativity, and reading comprehension. No big deal, the kind of stuff you've been doing in school all your life.

Now here's where Blake makes the experiment happen (or manipulates the independent variable in more technical terms). Blake randomly assigns you to one of three conditions while you complete the tests. The control group just takes the tests, "Thank you, here's your class credit, now go home." The Ignore TV group is told to complete the tests while the TV set is turned on, but to ignore the program and focus on the test. (Blake later gives these people a pop quiz on the program and, not too surprisingly, finds that they really didn't watch the show.) The third group, called Dual Task, takes the tests and is also told to watch the TV program and monitor it. (Blake gives a pop quiz to these folks to and they score much higher on it than the Ignore TV group.)

You can probably guess the outcomes. It is a nice little step ladder effect. First, the Dual Task group does the worst on the test. Next, the Ignore TV group does better (moderate effect) than the Dual Task group. Finally the control group does the best of all three (large effect versus Dual Task, moderate effect versus Ignore TV). So when we engage in that dual processing task (studying with TV) there is no facilitation effect, only interference as the two tasks disrupt each other.


It is apparent that media use typically does not have beneficial effects on our academic achievement. Among children and adults, across surveys and experiments (natural and laboratory), over many different academic indicators, TV use tends to reduce learning. And these effects are quite noticable.

Three factors contribute to this general negative effect. First, media use displaces study and practice time. Obviously every minute you give to TV is a minute taken away from something else. Second, TV use creates information overload in study settings. You simply cannot focus on TV and learning at the same time. And finally, it is also apparent that TV requires a different set of skills compared to academic skills. In other words the skills you need to be a good TV watcher are not the same skills you need to excel as a student.